多卡流水导出

.gitignore

@@ -128,3 +128,15 @@ Data/Pretrained
 Data/utils.py
 Experiment/checkpoint
 Experiment/log
+
+*.ckpt
+ckpts/unifolm_wma_dual.ckpt.prepared.pt
+trt_engines/video_backbone.onnx
+trt_engines/video_backbone.engine
+*.weight
+*.bias
+onnx__Mul*
+onnx__MatMul*
+onnx__Add_*
+Constant_*_attr__value
+*.0

ckpts/LICENSE

@@ -0,0 +1,439 @@
+Attribution-NonCommercial-ShareAlike 4.0 International
+
+Copyright (c) 2016-2025 HangZhou YuShu TECHNOLOGY CO.,LTD. ("Unitree Robotics")
+
+=======================================================================
+
+Creative Commons Corporation ("Creative Commons") is not a law firm and
+does not provide legal services or legal advice. Distribution of
+Creative Commons public licenses does not create a lawyer-client or
+other relationship. Creative Commons makes its licenses and related
+information available on an "as-is" basis. Creative Commons gives no
+warranties regarding its licenses, any material licensed under their
+terms and conditions, or any related information. Creative Commons
+disclaims all liability for damages resulting from their use to the
+fullest extent possible.
+
+Using Creative Commons Public Licenses
+
+Creative Commons public licenses provide a standard set of terms and
+conditions that creators and other rights holders may use to share
+original works of authorship and other material subject to copyright
+and certain other rights specified in the public license below. The
+following considerations are for informational purposes only, are not
+exhaustive, and do not form part of our licenses.
+
+     Considerations for licensors: Our public licenses are
+     intended for use by those authorized to give the public
+     permission to use material in ways otherwise restricted by
+     copyright and certain other rights. Our licenses are
+     irrevocable. Licensors should read and understand the terms
+     and conditions of the license they choose before applying it.
+     Licensors should also secure all rights necessary before
+     applying our licenses so that the public can reuse the
+     material as expected. Licensors should clearly mark any
+     material not subject to the license. This includes other CC-
+     licensed material, or material used under an exception or
+     limitation to copyright. More considerations for licensors:
+    wiki.creativecommons.org/Considerations_for_licensors
+
+     Considerations for the public: By using one of our public
+     licenses, a licensor grants the public permission to use the
+     licensed material under specified terms and conditions. If
+     the licensor's permission is not necessary for any reason--for
+     example, because of any applicable exception or limitation to
+     copyright--then that use is not regulated by the license. Our
+     licenses grant only permissions under copyright and certain
+     other rights that a licensor has authority to grant. Use of
+     the licensed material may still be restricted for other
+     reasons, including because others have copyright or other
+     rights in the material. A licensor may make special requests,
+     such as asking that all changes be marked or described.
+     Although not required by our licenses, you are encouraged to
+     respect those requests where reasonable. More considerations
+     for the public:
+    wiki.creativecommons.org/Considerations_for_licensees
+
+=======================================================================
+
+Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
+Public License
+
+By exercising the Licensed Rights (defined below), You accept and agree
+to be bound by the terms and conditions of this Creative Commons
+Attribution-NonCommercial-ShareAlike 4.0 International Public License
+("Public License"). To the extent this Public License may be
+interpreted as a contract, You are granted the Licensed Rights in
+consideration of Your acceptance of these terms and conditions, and the
+Licensor grants You such rights in consideration of benefits the
+Licensor receives from making the Licensed Material available under
+these terms and conditions.
+
+
+Section 1 -- Definitions.
+
+  a. Adapted Material means material subject to Copyright and Similar
+     Rights that is derived from or based upon the Licensed Material
+     and in which the Licensed Material is translated, altered,
+     arranged, transformed, or otherwise modified in a manner requiring
+     permission under the Copyright and Similar Rights held by the
+     Licensor. For purposes of this Public License, where the Licensed
+     Material is a musical work, performance, or sound recording,
+     Adapted Material is always produced where the Licensed Material is
+     synched in timed relation with a moving image.
+
+  b. Adapter's License means the license You apply to Your Copyright
+     and Similar Rights in Your contributions to Adapted Material in
+     accordance with the terms and conditions of this Public License.
+
+  c. BY-NC-SA Compatible License means a license listed at
+     creativecommons.org/compatiblelicenses, approved by Creative
+     Commons as essentially the equivalent of this Public License.
+
+  d. Copyright and Similar Rights means copyright and/or similar rights
+     closely related to copyright including, without limitation,
+     performance, broadcast, sound recording, and Sui Generis Database
+     Rights, without regard to how the rights are labeled or
+     categorized. For purposes of this Public License, the rights
+     specified in Section 2(b)(1)-(2) are not Copyright and Similar
+     Rights.
+
+  e. Effective Technological Measures means those measures that, in the
+     absence of proper authority, may not be circumvented under laws
+     fulfilling obligations under Article 11 of the WIPO Copyright
+     Treaty adopted on December 20, 1996, and/or similar international
+     agreements.
+
+  f. Exceptions and Limitations means fair use, fair dealing, and/or
+     any other exception or limitation to Copyright and Similar Rights
+     that applies to Your use of the Licensed Material.
+
+  g. License Elements means the license attributes listed in the name
+     of a Creative Commons Public License. The License Elements of this
+     Public License are Attribution, NonCommercial, and ShareAlike.
+
+  h. Licensed Material means the artistic or literary work, database,
+     or other material to which the Licensor applied this Public
+     License.
+
+  i. Licensed Rights means the rights granted to You subject to the
+     terms and conditions of this Public License, which are limited to
+     all Copyright and Similar Rights that apply to Your use of the
+     Licensed Material and that the Licensor has authority to license.
+
+  j. Licensor means the individual(s) or entity(ies) granting rights
+     under this Public License.
+
+  k. NonCommercial means not primarily intended for or directed towards
+     commercial advantage or monetary compensation. For purposes of
+     this Public License, the exchange of the Licensed Material for
+     other material subject to Copyright and Similar Rights by digital
+     file-sharing or similar means is NonCommercial provided there is
+     no payment of monetary compensation in connection with the
+     exchange.
+
+  l. Share means to provide material to the public by any means or
+     process that requires permission under the Licensed Rights, such
+     as reproduction, public display, public performance, distribution,
+     dissemination, communication, or importation, and to make material
+     available to the public including in ways that members of the
+     public may access the material from a place and at a time
+     individually chosen by them.
+
+  m. Sui Generis Database Rights means rights other than copyright
+     resulting from Directive 96/9/EC of the European Parliament and of
+     the Council of 11 March 1996 on the legal protection of databases,
+     as amended and/or succeeded, as well as other essentially
+     equivalent rights anywhere in the world.
+
+  n. You means the individual or entity exercising the Licensed Rights
+     under this Public License. Your has a corresponding meaning.
+
+
+Section 2 -- Scope.
+
+  a. License grant.
+
+       1. Subject to the terms and conditions of this Public License,
+          the Licensor hereby grants You a worldwide, royalty-free,
+          non-sublicensable, non-exclusive, irrevocable license to
+          exercise the Licensed Rights in the Licensed Material to:
+
+            a. reproduce and Share the Licensed Material, in whole or
+               in part, for NonCommercial purposes only; and
+
+            b. produce, reproduce, and Share Adapted Material for
+               NonCommercial purposes only.
+
+       2. Exceptions and Limitations. For the avoidance of doubt, where
+          Exceptions and Limitations apply to Your use, this Public
+          License does not apply, and You do not need to comply with
+          its terms and conditions.
+
+       3. Term. The term of this Public License is specified in Section
+          6(a).
+
+       4. Media and formats; technical modifications allowed. The
+          Licensor authorizes You to exercise the Licensed Rights in
+          all media and formats whether now known or hereafter created,
+          and to make technical modifications necessary to do so. The
+          Licensor waives and/or agrees not to assert any right or
+          authority to forbid You from making technical modifications
+          necessary to exercise the Licensed Rights, including
+          technical modifications necessary to circumvent Effective
+          Technological Measures. For purposes of this Public License,
+          simply making modifications authorized by this Section 2(a)
+          (4) never produces Adapted Material.
+
+       5. Downstream recipients.
+
+            a. Offer from the Licensor -- Licensed Material. Every
+               recipient of the Licensed Material automatically
+               receives an offer from the Licensor to exercise the
+               Licensed Rights under the terms and conditions of this
+               Public License.
+
+            b. Additional offer from the Licensor -- Adapted Material.
+               Every recipient of Adapted Material from You
+               automatically receives an offer from the Licensor to
+               exercise the Licensed Rights in the Adapted Material
+               under the conditions of the Adapter's License You apply.
+
+            c. No downstream restrictions. You may not offer or impose
+               any additional or different terms or conditions on, or
+               apply any Effective Technological Measures to, the
+               Licensed Material if doing so restricts exercise of the
+               Licensed Rights by any recipient of the Licensed
+               Material.
+
+       6. No endorsement. Nothing in this Public License constitutes or
+          may be construed as permission to assert or imply that You
+          are, or that Your use of the Licensed Material is, connected
+          with, or sponsored, endorsed, or granted official status by,
+          the Licensor or others designated to receive attribution as
+          provided in Section 3(a)(1)(A)(i).
+
+  b. Other rights.
+
+       1. Moral rights, such as the right of integrity, are not
+          licensed under this Public License, nor are publicity,
+          privacy, and/or other similar personality rights; however, to
+          the extent possible, the Licensor waives and/or agrees not to
+          assert any such rights held by the Licensor to the limited
+          extent necessary to allow You to exercise the Licensed
+          Rights, but not otherwise.
+
+       2. Patent and trademark rights are not licensed under this
+          Public License.
+
+       3. To the extent possible, the Licensor waives any right to
+          collect royalties from You for the exercise of the Licensed
+          Rights, whether directly or through a collecting society
+          under any voluntary or waivable statutory or compulsory
+          licensing scheme. In all other cases the Licensor expressly
+          reserves any right to collect such royalties, including when
+          the Licensed Material is used other than for NonCommercial
+          purposes.
+
+
+Section 3 -- License Conditions.
+
+Your exercise of the Licensed Rights is expressly made subject to the
+following conditions.
+
+  a. Attribution.
+
+       1. If You Share the Licensed Material (including in modified
+          form), You must:
+
+            a. retain the following if it is supplied by the Licensor
+               with the Licensed Material:
+
+                 i. identification of the creator(s) of the Licensed
+                    Material and any others designated to receive
+                    attribution, in any reasonable manner requested by
+                    the Licensor (including by pseudonym if
+                    designated);
+
+                ii. a copyright notice;
+
+               iii. a notice that refers to this Public License;
+
+                iv. a notice that refers to the disclaimer of
+                    warranties;
+
+                 v. a URI or hyperlink to the Licensed Material to the
+                    extent reasonably practicable;
+
+            b. indicate if You modified the Licensed Material and
+               retain an indication of any previous modifications; and
+
+            c. indicate the Licensed Material is licensed under this
+               Public License, and include the text of, or the URI or
+               hyperlink to, this Public License.
+
+       2. You may satisfy the conditions in Section 3(a)(1) in any
+          reasonable manner based on the medium, means, and context in
+          which You Share the Licensed Material. For example, it may be
+          reasonable to satisfy the conditions by providing a URI or
+          hyperlink to a resource that includes the required
+          information.
+       3. If requested by the Licensor, You must remove any of the
+          information required by Section 3(a)(1)(A) to the extent
+          reasonably practicable.
+
+  b. ShareAlike.
+
+     In addition to the conditions in Section 3(a), if You Share
+     Adapted Material You produce, the following conditions also apply.
+
+       1. The Adapter's License You apply must be a Creative Commons
+          license with the same License Elements, this version or
+          later, or a BY-NC-SA Compatible License.
+
+       2. You must include the text of, or the URI or hyperlink to, the
+          Adapter's License You apply. You may satisfy this condition
+          in any reasonable manner based on the medium, means, and
+          context in which You Share Adapted Material.
+
+       3. You may not offer or impose any additional or different terms
+          or conditions on, or apply any Effective Technological
+          Measures to, Adapted Material that restrict exercise of the
+          rights granted under the Adapter's License You apply.
+
+
+Section 4 -- Sui Generis Database Rights.
+
+Where the Licensed Rights include Sui Generis Database Rights that
+apply to Your use of the Licensed Material:
+
+  a. for the avoidance of doubt, Section 2(a)(1) grants You the right
+     to extract, reuse, reproduce, and Share all or a substantial
+     portion of the contents of the database for NonCommercial purposes
+     only;
+
+  b. if You include all or a substantial portion of the database
+     contents in a database in which You have Sui Generis Database
+     Rights, then the database in which You have Sui Generis Database
+     Rights (but not its individual contents) is Adapted Material,
+     including for purposes of Section 3(b); and
+
+  c. You must comply with the conditions in Section 3(a) if You Share
+     all or a substantial portion of the contents of the database.
+
+For the avoidance of doubt, this Section 4 supplements and does not
+replace Your obligations under this Public License where the Licensed
+Rights include other Copyright and Similar Rights.
+
+
+Section 5 -- Disclaimer of Warranties and Limitation of Liability.
+
+  a. UNLESS OTHERWISE SEPARATELY UNDERTAKEN BY THE LICENSOR, TO THE
+     EXTENT POSSIBLE, THE LICENSOR OFFERS THE LICENSED MATERIAL AS-IS
+     AND AS-AVAILABLE, AND MAKES NO REPRESENTATIONS OR WARRANTIES OF
+     ANY KIND CONCERNING THE LICENSED MATERIAL, WHETHER EXPRESS,
+     IMPLIED, STATUTORY, OR OTHER. THIS INCLUDES, WITHOUT LIMITATION,
+     WARRANTIES OF TITLE, MERCHANTABILITY, FITNESS FOR A PARTICULAR
+     PURPOSE, NON-INFRINGEMENT, ABSENCE OF LATENT OR OTHER DEFECTS,
+     ACCURACY, OR THE PRESENCE OR ABSENCE OF ERRORS, WHETHER OR NOT
+     KNOWN OR DISCOVERABLE. WHERE DISCLAIMERS OF WARRANTIES ARE NOT
+     ALLOWED IN FULL OR IN PART, THIS DISCLAIMER MAY NOT APPLY TO YOU.
+
+  b. TO THE EXTENT POSSIBLE, IN NO EVENT WILL THE LICENSOR BE LIABLE
+     TO YOU ON ANY LEGAL THEORY (INCLUDING, WITHOUT LIMITATION,
+     NEGLIGENCE) OR OTHERWISE FOR ANY DIRECT, SPECIAL, INDIRECT,
+     INCIDENTAL, CONSEQUENTIAL, PUNITIVE, EXEMPLARY, OR OTHER LOSSES,
+     COSTS, EXPENSES, OR DAMAGES ARISING OUT OF THIS PUBLIC LICENSE OR
+     USE OF THE LICENSED MATERIAL, EVEN IF THE LICENSOR HAS BEEN
+     ADVISED OF THE POSSIBILITY OF SUCH LOSSES, COSTS, EXPENSES, OR
+     DAMAGES. WHERE A LIMITATION OF LIABILITY IS NOT ALLOWED IN FULL OR
+     IN PART, THIS LIMITATION MAY NOT APPLY TO YOU.
+
+  c. The disclaimer of warranties and limitation of liability provided
+     above shall be interpreted in a manner that, to the extent
+     possible, most closely approximates an absolute disclaimer and
+     waiver of all liability.
+
+
+Section 6 -- Term and Termination.
+
+  a. This Public License applies for the term of the Copyright and
+     Similar Rights licensed here. However, if You fail to comply with
+     this Public License, then Your rights under this Public License
+     terminate automatically.
+
+  b. Where Your right to use the Licensed Material has terminated under
+     Section 6(a), it reinstates:
+
+       1. automatically as of the date the violation is cured, provided
+          it is cured within 30 days of Your discovery of the
+          violation; or
+
+       2. upon express reinstatement by the Licensor.
+
+     For the avoidance of doubt, this Section 6(b) does not affect any
+     right the Licensor may have to seek remedies for Your violations
+     of this Public License.
+
+  c. For the avoidance of doubt, the Licensor may also offer the
+     Licensed Material under separate terms or conditions or stop
+     distributing the Licensed Material at any time; however, doing so
+     will not terminate this Public License.
+
+  d. Sections 1, 5, 6, 7, and 8 survive termination of this Public
+     License.
+
+
+Section 7 -- Other Terms and Conditions.
+
+  a. The Licensor shall not be bound by any additional or different
+     terms or conditions communicated by You unless expressly agreed.
+
+  b. Any arrangements, understandings, or agreements regarding the
+     Licensed Material not stated herein are separate from and
+     independent of the terms and conditions of this Public License.
+
+
+Section 8 -- Interpretation.
+
+  a. For the avoidance of doubt, this Public License does not, and
+     shall not be interpreted to, reduce, limit, restrict, or impose
+     conditions on any use of the Licensed Material that could lawfully
+     be made without permission under this Public License.
+
+  b. To the extent possible, if any provision of this Public License is
+     deemed unenforceable, it shall be automatically reformed to the
+     minimum extent necessary to make it enforceable. If the provision
+     cannot be reformed, it shall be severed from this Public License
+     without affecting the enforceability of the remaining terms and
+     conditions.
+
+  c. No term or condition of this Public License will be waived and no
+     failure to comply consented to unless expressly agreed to by the
+     Licensor.
+
+  d. Nothing in this Public License constitutes or may be interpreted
+     as a limitation upon, or waiver of, any privileges and immunities
+     that apply to the Licensor or You, including from the legal
+     processes of any jurisdiction or authority.
+
+=======================================================================
+
+Creative Commons is not a party to its public
+licenses. Notwithstanding, Creative Commons may elect to apply one of
+its public licenses to material it publishes and in those instances
+will be considered the “Licensor.” The text of the Creative Commons
+public licenses is dedicated to the public domain under the CC0 Public
+Domain Dedication. Except for the limited purpose of indicating that
+material is shared under a Creative Commons public license or as
+otherwise permitted by the Creative Commons policies published at
+creativecommons.org/policies, Creative Commons does not authorize the
+use of the trademark "Creative Commons" or any other trademark or logo
+of Creative Commons without its prior written consent including,
+without limitation, in connection with any unauthorized modifications
+to any of its public licenses or any other arrangements,
+understandings, or agreements concerning use of licensed material. For
+the avoidance of doubt, this paragraph does not form part of the
+public licenses.
+
+Creative Commons may be contacted at creativecommons.org.

ckpts/README.md

@@ -0,0 +1,38 @@
+---
+tags:
+- robotics
+---
+
+# UnifoLM-WMA-0: A World-Model-Action (WMA) Framework under UnifoLM Family
+<p style="font-size: 1.2em;">
+    <a href="https://unigen-x.github.io/unifolm-world-model-action.github.io"><strong>Project Page</strong></a> | 
+    <a href="https://github.com/unitreerobotics/unifolm-world-model-action"><strong>Code</strong></a> |
+    <a href="https://huggingface.co/unitreerobotics/datasets"><strong>Dataset</strong></a> 
+  </p>
+<div align="center">
+  <div align="justify">
+    <b>UnifoLM-WMA-0</b> is Unitree‘s first open-source world-model–action architecture spanning multiple types of robotic embodiments, designed specifically for general-purpose robot learning. Its core component is a world-model capable of understanding the physical interactions between robots and the environments. This world-model provides two key functions: (a) <b>Simulation Engine</b> – operates as an interactive simulator to generate synthetic data for robot learning; (b) <b>Policy Enhancement</b> – connects with an action head and, by predicting future interaction processes with the world-model, further optimizes decision-making performance.
+  </div>
+</div>
+
+## 🦾 Real Robot Deployment
+| <img src="assets/real_z1_stackbox.gif" style="border:none;box-shadow:none;margin:0;padding:0;" /> | <img src="assets/real_dual_stackbox.gif" style="border:none;box-shadow:none;margin:0;padding:0;" /> |
+|:---:|:---:|
+| <img src="assets/real_cleanup_pencils.gif" style="border:none;box-shadow:none;margin:0;padding:0;" /> | <img src="assets/real_g1_pack_camera.gif" style="border:none;box-shadow:none;margin:0;padding:0;" /> |
+
+**Note: the top-right window shows the world model’s prediction of future environmental changes.**
+
+## License
+The model is released under the CC BY-NC-SA 4.0 license as found in the [LICENSE](https://huggingface.co/unitreerobotics/UnifoLM-WMA-0/blob/main/LICENSE). You are responsible for ensuring that your use of Unitree AI Models complies with all applicable laws.
+
+## Model Architecture
+![Demo](assets/world_model_interaction.gif)
+
+## Citation
+```
+@misc{unifolm-wma-0,
+  author = {Unitree},
+  title  = {UnifoLM-WMA-0: A World-Model-Action (WMA) Framework under UnifoLM Family},
+  year   = {2025},
+}
+```

configs/inference/world_model_interaction.yaml

@@ -222,7 +222,7 @@ data:
     test:
       target: unifolm_wma.data.wma_data.WMAData
       params:
-        data_dir: '/path/to/unifolm-world-model-action/examples/world_model_interaction_prompts'
+        data_dir: 'examples/world_model_interaction_prompts'
         video_length: ${model.params.wma_config.params.temporal_length}
         frame_stride: 2
         load_raw_resolution: True

psnr_score_for_challenge.py

@@ -0,0 +1,89 @@
+import os
+import glob
+import numpy as np
+import json
+from argparse import ArgumentParser, ArgumentDefaultsHelpFormatter
+from tqdm import tqdm
+from moviepy.video.io.VideoFileClip import VideoFileClip
+import PIL.Image
+
+
+def calculate_psnr(img1, img2):
+    mse = np.mean((img1.astype(np.float64) - img2.astype(np.float64)) ** 2)
+    if mse == 0:
+        return float('inf')
+    max_pixel = 255.0
+    psnr = 20 * np.log10(max_pixel / np.sqrt(mse))
+    return psnr
+
+
+def process_video_psnr(gt_path, pred_path):
+    try:
+        clip_gt = VideoFileClip(gt_path)
+        clip_pred = VideoFileClip(pred_path)
+
+        fps = min(clip_gt.fps, clip_pred.fps)
+        duration = min(clip_gt.duration, clip_pred.duration)
+
+        time_points = np.arange(0, duration, 1.0 / fps)
+
+        video_psnrs = []
+
+        for t in time_points:
+            frame_gt = clip_gt.get_frame(t)
+            frame_pred = clip_pred.get_frame(t)
+
+            img_gt = PIL.Image.fromarray(frame_gt).resize((256, 256), PIL.Image.Resampling.BILINEAR)
+            img_pred = PIL.Image.fromarray(frame_pred).resize((256, 256), PIL.Image.Resampling.BILINEAR)
+
+            psnr = calculate_psnr(np.array(img_gt), np.array(img_pred))
+            video_psnrs.append(psnr)
+
+        clip_gt.close()
+        clip_pred.close()
+
+        return np.mean(video_psnrs) if video_psnrs else 0.0
+
+    except Exception as e:
+        print(f"Error processing {os.path.basename(gt_path)}: {e}")
+        return None
+
+
+def main():
+    parser = ArgumentParser(formatter_class=ArgumentDefaultsHelpFormatter)
+    parser.add_argument('--gt_video', type=str, required=True, help='path to reference videos')
+    parser.add_argument('--pred_video', type=str, required=True, help='path to pred videos')
+    parser.add_argument('--output_file', type=str, default=None, help='path to output file')
+    args = parser.parse_args()
+
+    if not os.path.exists(args.gt_video):
+        print(f"Error: GT video not found at {args.gt_video}")
+        return
+    if not os.path.exists(args.pred_video):
+        print(f"Error: Pred video not found at {args.pred_video}")
+        return
+
+    print(f"Comparing:\nRef: {args.gt_video}\nPred: {args.pred_video}")
+
+    v_psnr = process_video_psnr(args.gt_video, args.pred_video)
+
+    if v_psnr is not None:
+        print("-" * 30)
+        print(f"Video PSNR: {v_psnr:.4f} dB")
+        print("-" * 30)
+
+        if args.output_file:
+            result = {
+                "gt_video": args.gt_video,
+                "pred_video": args.pred_video,
+                "psnr": v_psnr
+            }
+            with open(args.output_file, 'w') as f:
+                json.dump(result, f, indent=4)
+            print(f"Result saved to {args.output_file}")
+    else:
+        print("Failed to calculate PSNR.")
+
+
+if __name__ == '__main__':
+    main()

pyproject.toml

@@ -19,13 +19,13 @@ dependencies = [
     "pytorch-lightning==1.9.3",
     "pyyaml==6.0",
     "setuptools==65.6.3",
-    "torch==2.3.1",
+    #"torch==2.3.1",
-    "torchvision==0.18.1",
+    #"torchvision==0.18.1",
     "tqdm==4.66.5",
     "transformers==4.40.1",
     "moviepy==1.0.3",
     "av==12.3.0",
-    "xformers==0.0.27",
+    #"xformers==0.0.27",
     "gradio==4.39.0",
     "timm==0.9.10",
     "scikit-learn==1.5.1",

scripts/evaluation/analyze_backbone_blocks.py

@@ -0,0 +1,148 @@
+import argparse
+from pathlib import Path
+
+import numpy as np
+import pandas as pd
+
+
+MID_STEP_MIN = 15
+MID_STEP_MAX = 35
+TAIL_STEP_MIN = 40
+TAIL_STEP_MAX = 49
+
+
+def describe(series: pd.Series) -> dict[str, float]:
+    numeric = pd.to_numeric(series, errors='coerce').dropna()
+    if numeric.empty:
+        return {'count': 0.0, 'mean': np.nan, 'median': np.nan, 'p90': np.nan}
+    return {
+        'count': float(numeric.count()),
+        'mean': float(numeric.mean()),
+        'median': float(numeric.median()),
+        'p90': float(numeric.quantile(0.90)),
+    }
+
+
+def make_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="Aggregate backbone block profiling results.")
+    parser.add_argument("--input_csv",
+                        type=str,
+                        required=True,
+                        help="Path to backbone_block_log.csv")
+    parser.add_argument("--output_dir",
+                        type=str,
+                        default=None,
+                        help="Directory to store summaries; defaults to input parent.")
+    return parser
+
+
+def main() -> None:
+    args = make_parser().parse_args()
+    input_path = Path(args.input_csv)
+    if not input_path.exists():
+        raise FileNotFoundError(f"Missing file: {input_path}")
+
+    output_dir = Path(args.output_dir) if args.output_dir else input_path.parent / "analysis"
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+    df = pd.read_csv(input_path)
+    numeric_cols = [
+        'forward_time_ms',
+        'l2_delta_vs_prev',
+        'rel_l2_delta_vs_prev',
+        'cosine_vs_prev',
+        'l2_delta_vs_full50',
+        'cosine_vs_full50',
+    ]
+    for col in numeric_cols:
+        df[col] = pd.to_numeric(df[col], errors='coerce')
+
+    step_aggregate = df.groupby(
+        ['block_stage', 'block_name', 'block_index', 'step'])[numeric_cols].mean().reset_index()
+    step_aggregate.to_csv(output_dir / 'block_step_aggregate.csv', index=False)
+
+    block_summary = df.groupby(['block_stage', 'block_name', 'block_index']).agg(
+        mean_forward_time_ms=('forward_time_ms', 'mean'),
+        mean_rel_l2_delta_vs_prev=('rel_l2_delta_vs_prev', 'mean'),
+        mean_cosine_vs_prev=('cosine_vs_prev', 'mean'),
+        mean_l2_delta_vs_full50=('l2_delta_vs_full50', 'mean'),
+        mean_cosine_vs_full50=('cosine_vs_full50', 'mean'),
+    ).reset_index()
+
+    mid_df = df[(df['step'] >= MID_STEP_MIN) & (df['step'] <= MID_STEP_MAX)]
+    tail_df = df[(df['step'] >= TAIL_STEP_MIN) & (df['step'] <= TAIL_STEP_MAX)]
+    block_summary = block_summary.merge(
+        mid_df.groupby(['block_stage', 'block_name', 'block_index']).agg(
+            mid_mean_rel_l2_delta_vs_prev=('rel_l2_delta_vs_prev', 'mean'),
+            mid_mean_cosine_vs_prev=('cosine_vs_prev', 'mean'),
+            mid_mean_forward_time_ms=('forward_time_ms', 'mean'),
+        ).reset_index(),
+        on=['block_stage', 'block_name', 'block_index'],
+        how='left',
+    )
+    block_summary = block_summary.merge(
+        tail_df.groupby(['block_stage', 'block_name', 'block_index']).agg(
+            tail_mean_rel_l2_delta_vs_prev=('rel_l2_delta_vs_prev', 'mean'),
+            tail_mean_cosine_vs_prev=('cosine_vs_prev', 'mean'),
+            tail_mean_forward_time_ms=('forward_time_ms', 'mean'),
+        ).reset_index(),
+        on=['block_stage', 'block_name', 'block_index'],
+        how='left',
+    )
+    block_summary = block_summary.sort_values(
+        ['tail_mean_rel_l2_delta_vs_prev', 'mean_forward_time_ms'],
+        ascending=[True, False])
+    block_summary.to_csv(output_dir / 'block_summary.csv', index=False)
+
+    stage_summary = []
+    for stage_name, stage_df in [('all', df), ('mid', mid_df), ('tail', tail_df)]:
+        grouped = stage_df.groupby('block_stage').agg(
+            mean_forward_time_ms=('forward_time_ms', 'mean'),
+            mean_rel_l2_delta_vs_prev=('rel_l2_delta_vs_prev', 'mean'),
+            mean_cosine_vs_prev=('cosine_vs_prev', 'mean'),
+            mean_l2_delta_vs_full50=('l2_delta_vs_full50', 'mean'),
+            mean_cosine_vs_full50=('cosine_vs_full50', 'mean'),
+        ).reset_index()
+        grouped.insert(0, 'window', stage_name)
+        stage_summary.append(grouped)
+    stage_summary_df = pd.concat(stage_summary, ignore_index=True)
+    stage_summary_df.to_csv(output_dir / 'stage_summary.csv', index=False)
+
+    best_cache_candidates = block_summary.sort_values(
+        ['tail_mean_rel_l2_delta_vs_prev', 'mean_forward_time_ms'],
+        ascending=[True, False]).head(10)
+
+    lines = [
+        "# Backbone Block Profiling Report",
+        "",
+        "## Dataset Overview",
+        "",
+        f"- Rows: {len(df)}",
+        f"- Blocks: {df['block_name'].nunique()}",
+        f"- Steps: {int(df['step'].min())}-{int(df['step'].max())}",
+        f"- Pass types: {', '.join(sorted(df['pass_type'].dropna().unique()))}",
+        "",
+        "## Block Timing",
+        "",
+        f"- `forward_time_ms`: mean={describe(df['forward_time_ms'])['mean']:.4f}, median={describe(df['forward_time_ms'])['median']:.4f}, p90={describe(df['forward_time_ms'])['p90']:.4f}",
+        "",
+        "## Stability",
+        "",
+        f"- `rel_l2_delta_vs_prev`: mean={describe(df['rel_l2_delta_vs_prev'])['mean']:.6f}, median={describe(df['rel_l2_delta_vs_prev'])['median']:.6f}, p90={describe(df['rel_l2_delta_vs_prev'])['p90']:.6f}",
+        f"- `cosine_vs_prev`: mean={describe(df['cosine_vs_prev'])['mean']:.6f}, median={describe(df['cosine_vs_prev'])['median']:.6f}, p90={describe(df['cosine_vs_prev'])['p90']:.6f}",
+        "",
+        "## Top Cache Candidates",
+        "",
+    ]
+    for _, row in best_cache_candidates.iterrows():
+        lines.append(
+            f"- `{row['block_name']}` ({row['block_stage']}): tail_rel_l2={row['tail_mean_rel_l2_delta_vs_prev']:.6f}, mean_forward_time_ms={row['mean_forward_time_ms']:.4f}"
+        )
+
+    (output_dir / 'backbone_profile_report.md').write_text("\n".join(lines),
+                                                           encoding='utf-8')
+
+
+if __name__ == "__main__":
+    main()

scripts/evaluation/analyze_metrics.py

@@ -0,0 +1,611 @@
+import argparse
+import json
+import os
+from pathlib import Path
+
+import numpy as np
+import pandas as pd
+
+try:
+    import matplotlib.pyplot as plt
+except ModuleNotFoundError:
+    plt = None
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Analyze DDIM convergence logs for the five research directions."
+    )
+    parser.add_argument(
+        "--input_dir",
+        type=str,
+        required=True,
+        help="Directory containing stepwise_log.csv, sample_summary.csv, and round_summary.csv.",
+    )
+    parser.add_argument(
+        "--output_dir",
+        type=str,
+        default=None,
+        help="Directory to write analysis outputs. Defaults to <input_dir>/analysis.",
+    )
+    parser.add_argument(
+        "--mid_step_start",
+        type=int,
+        default=15,
+        help="Start of the step window used for cross-step latent-delta analysis.",
+    )
+    parser.add_argument(
+        "--mid_step_end",
+        type=int,
+        default=35,
+        help="End of the step window used for cross-step latent-delta analysis.",
+    )
+    parser.add_argument(
+        "--full50_step_cap",
+        type=int,
+        default=50,
+        help="Upper bound used when summarizing step-based metrics.",
+    )
+    parser.add_argument(
+        "--stability_threshold",
+        type=float,
+        default=None,
+        help="Optional threshold for offline first_stable_step analysis. If unset, stable-step metrics are not computed.",
+    )
+    parser.add_argument(
+        "--stability_window",
+        type=int,
+        default=3,
+        help="Consecutive-step window for offline first_stable_step analysis.",
+    )
+    return parser.parse_args()
+
+
+def safe_float(value) -> float:
+    if pd.isna(value):
+        return float("nan")
+    return float(value)
+
+
+def describe_series(series: pd.Series) -> dict[str, float]:
+    series = pd.to_numeric(series, errors="coerce").dropna()
+    if series.empty:
+        return {
+            "count": 0.0,
+            "mean": float("nan"),
+            "std": float("nan"),
+            "min": float("nan"),
+            "p25": float("nan"),
+            "median": float("nan"),
+            "p75": float("nan"),
+            "p90": float("nan"),
+            "p95": float("nan"),
+            "max": float("nan"),
+        }
+    return {
+        "count": float(series.count()),
+        "mean": float(series.mean()),
+        "std": float(series.std(ddof=0)),
+        "min": float(series.min()),
+        "p25": float(series.quantile(0.25)),
+        "median": float(series.median()),
+        "p75": float(series.quantile(0.75)),
+        "p90": float(series.quantile(0.90)),
+        "p95": float(series.quantile(0.95)),
+        "max": float(series.max()),
+    }
+
+
+def ensure_columns(frame: pd.DataFrame, required: list[str], frame_name: str) -> None:
+    missing = [column for column in required if column not in frame.columns]
+    if missing:
+        raise ValueError(f"{frame_name} is missing required columns: {missing}")
+
+
+def first_consecutive_below(values: pd.Series, threshold: float,
+                            window: int) -> float:
+    cleaned = pd.to_numeric(values, errors="coerce").tolist()
+    if window <= 0 or len(cleaned) < window:
+        return float("nan")
+    for start in range(len(cleaned) - window + 1):
+        chunk = cleaned[start:start + window]
+        if all(pd.notna(value) and value < threshold for value in chunk):
+            return float(start + 1)
+    return float("nan")
+
+
+def load_tables(input_dir: Path) -> tuple[pd.DataFrame, pd.DataFrame, pd.DataFrame]:
+    stepwise_path = input_dir / "stepwise_log.csv"
+    sample_summary_path = input_dir / "sample_summary.csv"
+    round_summary_path = input_dir / "round_summary.csv"
+
+    if not stepwise_path.exists():
+        raise FileNotFoundError(f"Missing file: {stepwise_path}")
+    if not sample_summary_path.exists():
+        raise FileNotFoundError(f"Missing file: {sample_summary_path}")
+    if not round_summary_path.exists():
+        raise FileNotFoundError(f"Missing file: {round_summary_path}")
+
+    stepwise_df = pd.read_csv(stepwise_path)
+    sample_summary_df = pd.read_csv(sample_summary_path)
+    round_summary_df = pd.read_csv(round_summary_path)
+
+    ensure_columns(
+        stepwise_df,
+        [
+            "sample_id",
+            "scene",
+            "pass_type",
+            "round_id",
+            "step",
+            "latent_delta",
+            "action_delta",
+            "state_delta",
+            "action_cosine_vs_full50",
+            "state_cosine_vs_full50",
+            "latent_l2_vs_full50",
+        ],
+        "stepwise_log.csv",
+    )
+    ensure_columns(
+        sample_summary_df,
+        [
+            "sample_id",
+            "scene",
+            "pass_type",
+            "pass_total_time_s",
+            "action_first_stable_step",
+            "state_first_stable_step",
+            "latent_first_stable_step",
+            "action_vs_full50_90pct_step",
+            "action_vs_full50_95pct_step",
+            "oracle_budget_action",
+            "oracle_budget_state",
+        ],
+        "sample_summary.csv",
+    )
+    ensure_columns(
+        round_summary_df,
+        [
+            "sample_id",
+            "scene",
+            "round_id",
+            "policy_pass_total_time_s",
+            "world_model_pass_total_time_s",
+            "round_total_time_s",
+            "latent_init_dist_to_prev_round",
+            "action_drift_vs_prev_round",
+        ],
+        "round_summary.csv",
+    )
+    return stepwise_df, sample_summary_df, round_summary_df
+
+
+def aggregate_stepwise(stepwise_df: pd.DataFrame) -> pd.DataFrame:
+    metric_columns = [
+        "step_time_s",
+        "latent_delta",
+        "action_delta",
+        "state_delta",
+        "action_cosine_vs_full50",
+        "state_cosine_vs_full50",
+        "latent_l2_vs_full50",
+    ]
+    stepwise_df = stepwise_df.copy()
+    for column in metric_columns:
+        stepwise_df[column] = pd.to_numeric(stepwise_df[column], errors="coerce")
+    grouped = stepwise_df.groupby(["pass_type", "step"])[metric_columns]
+    return grouped.agg(["mean", "median", "std"]).reset_index()
+
+
+def aggregate_scene_summary(sample_summary_df: pd.DataFrame) -> pd.DataFrame:
+    metrics = [
+        "pass_total_time_s",
+        "action_first_stable_step",
+        "state_first_stable_step",
+        "latent_first_stable_step",
+        "action_vs_full50_90pct_step",
+        "action_vs_full50_95pct_step",
+        "oracle_budget_action",
+        "oracle_budget_state",
+        "round_total_time_s",
+        "policy_pass_total_time_s",
+        "world_model_pass_total_time_s",
+        "latent_init_dist_to_prev_round",
+        "action_drift_vs_prev_round",
+    ]
+    sample_summary_df = sample_summary_df.copy()
+    for column in metrics:
+        sample_summary_df[column] = pd.to_numeric(sample_summary_df[column],
+                                                  errors="coerce")
+    grouped = sample_summary_df.groupby(["scene", "pass_type"])[metrics]
+    return grouped.agg(["mean", "median", "std"]).reset_index()
+
+
+def compute_stability_summary(stepwise_df: pd.DataFrame, threshold: float,
+                              window: int) -> pd.DataFrame:
+    rows = []
+    grouped = stepwise_df.sort_values("step").groupby(
+        ["sample_id", "scene", "pass_type"])
+    for (sample_id, scene, pass_type), group in grouped:
+        rows.append({
+            "sample_id": sample_id,
+            "scene": scene,
+            "pass_type": pass_type,
+            "action_first_stable_step": first_consecutive_below(
+                group["action_delta"], threshold, window),
+            "state_first_stable_step": first_consecutive_below(
+                group["state_delta"], threshold, window),
+            "latent_first_stable_step": first_consecutive_below(
+                group["latent_delta"], threshold, window),
+        })
+    return pd.DataFrame(rows)
+
+
+def compute_direction_summary(stepwise_df: pd.DataFrame,
+                              sample_summary_df: pd.DataFrame,
+                              round_summary_df: pd.DataFrame,
+                              mid_step_start: int,
+                              mid_step_end: int,
+                              stability_threshold: float | None,
+                              stability_window: int) -> dict:
+    policy_summary = sample_summary_df[
+        sample_summary_df["pass_type"] == "policy"].copy()
+    world_summary = sample_summary_df[
+        sample_summary_df["pass_type"] == "world_model"].copy()
+
+    latent_mid = stepwise_df[
+        (stepwise_df["step"] >= mid_step_start)
+        & (stepwise_df["step"] <= mid_step_end)].copy()
+
+    action_before_latent = policy_summary[[
+        "action_first_stable_step",
+        "latent_first_stable_step",
+        "action_vs_full50_95pct_step",
+    ]].dropna()
+    if action_before_latent.empty:
+        action_first_beats_latent = float("nan")
+        action_95_beats_latent = float("nan")
+    else:
+        action_first_beats_latent = float(
+            (action_before_latent["action_first_stable_step"] <
+             action_before_latent["latent_first_stable_step"]).mean())
+        action_95_beats_latent = float(
+            (action_before_latent["action_vs_full50_95pct_step"] <
+             action_before_latent["latent_first_stable_step"]).mean())
+
+    direction_summary = {
+        "analysis_config": {
+            "stability_threshold": stability_threshold,
+            "stability_window": int(stability_window),
+            "mid_step_window": [int(mid_step_start), int(mid_step_end)],
+        },
+        "dataset_overview": {
+            "num_step_rows": int(len(stepwise_df)),
+            "num_sample_rows": int(len(sample_summary_df)),
+            "num_round_rows": int(len(round_summary_df)),
+            "num_unique_samples": int(sample_summary_df["sample_id"].nunique()),
+            "scenes": sorted(sample_summary_df["scene"].dropna().unique().tolist()),
+            "pass_types": sorted(sample_summary_df["pass_type"].dropna().unique().tolist()),
+        },
+        "direction_original_early_stop": {
+            "latent_first_stable_step_policy":
+            describe_series(policy_summary["latent_first_stable_step"]),
+            "latent_first_stable_step_world_model":
+            describe_series(world_summary["latent_first_stable_step"]),
+            "latent_l2_vs_full50":
+            describe_series(stepwise_df["latent_l2_vs_full50"]),
+        },
+        "direction_c_action_converges_first": {
+            "action_first_stable_step":
+            describe_series(policy_summary["action_first_stable_step"]),
+            "latent_first_stable_step":
+            describe_series(policy_summary["latent_first_stable_step"]),
+            "action_vs_full50_95pct_step":
+            describe_series(policy_summary["action_vs_full50_95pct_step"]),
+            "share_action_first_stable_before_latent":
+            action_first_beats_latent,
+            "share_action_95pct_before_latent":
+            action_95_beats_latent,
+        },
+        "direction_d_cross_step_similarity": {
+            "latent_delta_mid_steps":
+            describe_series(latent_mid["latent_delta"]),
+            "action_delta_mid_steps":
+            describe_series(latent_mid["action_delta"]),
+            "state_delta_mid_steps":
+            describe_series(latent_mid["state_delta"]),
+        },
+        "direction_a_budget_heterogeneity": {
+            "oracle_budget_action":
+            describe_series(policy_summary["oracle_budget_action"]),
+            "oracle_budget_state":
+            describe_series(world_summary["oracle_budget_state"]),
+            "oracle_budget_action_by_scene": {
+                scene: describe_series(group["oracle_budget_action"])
+                for scene, group in policy_summary.groupby("scene")
+            },
+        },
+        "direction_b_round_reuse": {
+            "latent_init_dist_to_prev_round":
+            describe_series(round_summary_df["latent_init_dist_to_prev_round"]),
+            "action_drift_vs_prev_round":
+            describe_series(round_summary_df["action_drift_vs_prev_round"]),
+            "round_total_time_s":
+            describe_series(round_summary_df["round_total_time_s"]),
+            "policy_pass_total_time_s":
+            describe_series(round_summary_df["policy_pass_total_time_s"]),
+            "world_model_pass_total_time_s":
+            describe_series(round_summary_df["world_model_pass_total_time_s"]),
+        },
+    }
+    return direction_summary
+
+
+def save_json(path: Path, payload: dict) -> None:
+    with open(path, "w", encoding="utf-8") as file:
+        json.dump(payload, file, indent=2, ensure_ascii=False)
+
+
+def write_markdown_report(path: Path, direction_summary: dict) -> None:
+    lines = [
+        "# DDIM Analysis Report",
+        "",
+        "## Analysis Config",
+        "",
+    ]
+    config = direction_summary["analysis_config"]
+    lines.extend([
+        f"- Stability threshold: {config['stability_threshold']}",
+        f"- Stability window: {config['stability_window']}",
+        f"- Mid-step window: {config['mid_step_window'][0]}-{config['mid_step_window'][1]}",
+        "",
+        "## Dataset Overview",
+        "",
+    ])
+    overview = direction_summary["dataset_overview"]
+    lines.extend([
+        f"- Unique samples: {overview['num_unique_samples']}",
+        f"- Step rows: {overview['num_step_rows']}",
+        f"- Sample rows: {overview['num_sample_rows']}",
+        f"- Round rows: {overview['num_round_rows']}",
+        f"- Scenes: {', '.join(overview['scenes'])}",
+        f"- Pass types: {', '.join(overview['pass_types'])}",
+        "",
+        "## Five Directions",
+        "",
+    ])
+
+    def append_stats(title: str, payload: dict) -> None:
+        lines.append(f"### {title}")
+        lines.append("")
+        for key, value in payload.items():
+            if isinstance(value, dict):
+                if {"median", "mean", "p90"} <= set(value.keys()):
+                    lines.append(
+                        f"- `{key}`: mean={value['mean']:.4f}, median={value['median']:.4f}, p90={value['p90']:.4f}, p95={value['p95']:.4f}"
+                    )
+                else:
+                    lines.append(f"- `{key}`:")
+                    for sub_key, sub_value in value.items():
+                        if isinstance(sub_value, dict) and {"median", "mean"} <= set(
+                                sub_value.keys()):
+                            lines.append(
+                                f"  - `{sub_key}`: mean={sub_value['mean']:.4f}, median={sub_value['median']:.4f}, std={sub_value['std']:.4f}"
+                            )
+                        else:
+                            lines.append(f"  - `{sub_key}`: {sub_value}")
+            else:
+                if isinstance(value, float):
+                    lines.append(f"- `{key}`: {value:.4f}")
+                else:
+                    lines.append(f"- `{key}`: {value}")
+        lines.append("")
+
+    append_stats("Original Early Stop",
+                 direction_summary["direction_original_early_stop"])
+    append_stats("Direction C: Action Converges First",
+                 direction_summary["direction_c_action_converges_first"])
+    append_stats("Direction D: Cross-Step Similarity",
+                 direction_summary["direction_d_cross_step_similarity"])
+    append_stats("Direction A: Budget Heterogeneity",
+                 direction_summary["direction_a_budget_heterogeneity"])
+    append_stats("Direction B: Round Reuse",
+                 direction_summary["direction_b_round_reuse"])
+
+    with open(path, "w", encoding="utf-8") as file:
+        file.write("\n".join(lines) + "\n")
+
+
+def save_convergence_plot(stepwise_df: pd.DataFrame, output_path: Path) -> None:
+    if plt is None:
+        return
+    fig, axes = plt.subplots(2, 2, figsize=(14, 10))
+    pass_types = sorted(stepwise_df["pass_type"].dropna().unique().tolist())
+    for pass_type in pass_types:
+        subset = stepwise_df[stepwise_df["pass_type"] == pass_type]
+        grouped = subset.groupby("step").mean(numeric_only=True).reset_index()
+        axes[0, 0].plot(grouped["step"],
+                        grouped["latent_delta"],
+                        label=pass_type)
+        axes[0, 1].plot(grouped["step"],
+                        grouped["action_cosine_vs_full50"],
+                        label=pass_type)
+        axes[1, 0].plot(grouped["step"],
+                        grouped["state_cosine_vs_full50"],
+                        label=pass_type)
+        axes[1, 1].plot(grouped["step"],
+                        grouped["latent_l2_vs_full50"],
+                        label=pass_type)
+
+    axes[0, 0].set_title("Mean Latent Delta")
+    axes[0, 1].set_title("Mean Action Cosine vs Full50")
+    axes[1, 0].set_title("Mean State Cosine vs Full50")
+    axes[1, 1].set_title("Mean Latent L2 vs Full50")
+    for axis in axes.flatten():
+        axis.set_xlabel("Step")
+        axis.grid(alpha=0.3)
+        axis.legend()
+    fig.tight_layout()
+    fig.savefig(output_path, dpi=200, bbox_inches="tight")
+    plt.close(fig)
+
+
+def save_budget_distribution_plot(sample_summary_df: pd.DataFrame,
+                                  output_path: Path) -> None:
+    if plt is None:
+        return
+    policy_summary = sample_summary_df[
+        sample_summary_df["pass_type"] == "policy"].copy()
+    scenes = sorted(policy_summary["scene"].dropna().unique().tolist())
+    fig, axes = plt.subplots(1, 2, figsize=(14, 5))
+
+    for scene in scenes:
+        subset = policy_summary[policy_summary["scene"] == scene]
+        axes[0].hist(subset["oracle_budget_action"].dropna(),
+                     bins=15,
+                     alpha=0.5,
+                     label=scene)
+    axes[0].set_title("Oracle Budget Action Distribution")
+    axes[0].set_xlabel("Step")
+    axes[0].set_ylabel("Count")
+    axes[0].legend()
+    axes[0].grid(alpha=0.3)
+
+    boxplot_data = [
+        policy_summary[policy_summary["scene"] == scene]["oracle_budget_action"].dropna()
+        for scene in scenes
+    ]
+    if scenes and any(len(values) > 0 for values in boxplot_data):
+        axes[1].boxplot(boxplot_data, labels=scenes, showfliers=False)
+    axes[1].set_title("Oracle Budget Action by Scene")
+    axes[1].set_ylabel("Step")
+    axes[1].tick_params(axis="x", rotation=20)
+    axes[1].grid(alpha=0.3)
+
+    fig.tight_layout()
+    fig.savefig(output_path, dpi=200, bbox_inches="tight")
+    plt.close(fig)
+
+
+def save_midstep_delta_plot(stepwise_df: pd.DataFrame, output_path: Path,
+                            mid_step_start: int, mid_step_end: int) -> None:
+    if plt is None:
+        return
+    subset = stepwise_df[(stepwise_df["step"] >= mid_step_start)
+                         & (stepwise_df["step"] <= mid_step_end)].copy()
+    steps = sorted(subset["step"].dropna().unique().tolist())
+    data = [subset[subset["step"] == step]["latent_delta"].dropna() for step in steps]
+
+    fig, ax = plt.subplots(figsize=(14, 5))
+    if steps and any(len(values) > 0 for values in data):
+        ax.boxplot(data, labels=steps, showfliers=False)
+    ax.set_title(
+        f"Latent Delta Distribution for Steps {mid_step_start}-{mid_step_end}")
+    ax.set_xlabel("Step")
+    ax.set_ylabel("latent_delta")
+    ax.grid(alpha=0.3)
+    fig.tight_layout()
+    fig.savefig(output_path, dpi=200, bbox_inches="tight")
+    plt.close(fig)
+
+
+def save_round_reuse_plot(round_summary_df: pd.DataFrame, output_path: Path) -> None:
+    if plt is None:
+        return
+    fig, axes = plt.subplots(1, 2, figsize=(14, 5))
+
+    subset = round_summary_df.dropna(
+        subset=["latent_init_dist_to_prev_round", "action_drift_vs_prev_round"])
+    if not subset.empty:
+        for scene, group in subset.groupby("scene"):
+            axes[0].scatter(group["latent_init_dist_to_prev_round"],
+                            group["action_drift_vs_prev_round"],
+                            alpha=0.6,
+                            label=scene)
+        axes[0].legend()
+    axes[0].set_title("Round Reuse: Latent Init Dist vs Action Drift")
+    axes[0].set_xlabel("latent_init_dist_to_prev_round")
+    axes[0].set_ylabel("action_drift_vs_prev_round")
+    axes[0].grid(alpha=0.3)
+
+    scene_groups = []
+    scene_labels = []
+    for scene, group in round_summary_df.groupby("scene"):
+        values = group["round_total_time_s"].dropna()
+        if len(values) > 0:
+            scene_groups.append(values)
+            scene_labels.append(scene)
+    if scene_groups:
+        axes[1].boxplot(scene_groups, labels=scene_labels, showfliers=False)
+    axes[1].set_title("Round Total Time by Scene")
+    axes[1].set_ylabel("Seconds")
+    axes[1].tick_params(axis="x", rotation=20)
+    axes[1].grid(alpha=0.3)
+
+    fig.tight_layout()
+    fig.savefig(output_path, dpi=200, bbox_inches="tight")
+    plt.close(fig)
+
+
+def main() -> None:
+    args = parse_args()
+    input_dir = Path(args.input_dir)
+    output_dir = Path(args.output_dir) if args.output_dir else input_dir / "analysis"
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+    stepwise_df, sample_summary_df, round_summary_df = load_tables(input_dir)
+
+    stability_summary_df = None
+    sample_summary_for_analysis = sample_summary_df.copy()
+    if args.stability_threshold is not None:
+        stability_summary_df = compute_stability_summary(stepwise_df,
+                                                         args.stability_threshold,
+                                                         args.stability_window)
+        sample_summary_for_analysis = sample_summary_for_analysis.drop(
+            columns=[
+                "action_first_stable_step",
+                "state_first_stable_step",
+                "latent_first_stable_step",
+            ],
+            errors="ignore",
+        ).merge(stability_summary_df,
+                on=["sample_id", "scene", "pass_type"],
+                how="left")
+
+    step_aggregate_df = aggregate_stepwise(stepwise_df)
+    scene_summary_df = aggregate_scene_summary(sample_summary_for_analysis)
+    direction_summary = compute_direction_summary(
+        stepwise_df=stepwise_df,
+        sample_summary_df=sample_summary_for_analysis,
+        round_summary_df=round_summary_df,
+        mid_step_start=args.mid_step_start,
+        mid_step_end=args.mid_step_end,
+        stability_threshold=args.stability_threshold,
+        stability_window=args.stability_window,
+    )
+
+    step_aggregate_df.to_csv(output_dir / "step_aggregate.csv", index=False)
+    scene_summary_df.to_csv(output_dir / "scene_summary.csv", index=False)
+    if stability_summary_df is not None:
+        stability_summary_df.to_csv(output_dir / "stability_summary.csv",
+                                    index=False)
+    save_json(output_dir / "direction_summary.json", direction_summary)
+    write_markdown_report(output_dir / "analysis_report.md", direction_summary)
+
+    if plt is not None:
+        save_convergence_plot(stepwise_df, output_dir / "convergence_curves.png")
+        save_budget_distribution_plot(sample_summary_df,
+                                      output_dir / "budget_distribution.png")
+        save_midstep_delta_plot(stepwise_df,
+                                output_dir / "midstep_latent_delta.png",
+                                args.mid_step_start, args.mid_step_end)
+        save_round_reuse_plot(round_summary_df, output_dir / "round_reuse.png")
+    else:
+        print("matplotlib not installed; skipped PNG plots.")
+
+    print(f"Analysis written to: {output_dir}")
+
+
+if __name__ == "__main__":
+    main()

scripts/evaluation/simulate_sparse_head_execution.py

@@ -0,0 +1,380 @@
+import argparse
+import json
+from pathlib import Path
+
+import numpy as np
+import pandas as pd
+
+
+DEFAULT_SCHEMES = {
+    "dense": list(range(50)),
+    "sparse_10": list(range(0, 50, 5)),
+    "sparse_5": list(range(0, 50, 10)),
+    "tail_heavy_10": [0, 5, 10, 20, 30, 38, 43, 46, 48, 49],
+    "tail_only_6": [40, 43, 46, 47, 48, 49],
+    "sparse_8": [0, 7, 14, 21, 28, 35, 42, 49],
+    "sparse_4": [0, 16, 32, 49],
+    "tail_only_4": [40, 43, 46, 49],
+    "tail_heavy_6": [0, 32, 40, 44, 47, 49],
+    "tail_heavy_8": [0, 10, 20, 30, 38, 43, 46, 49],
+}
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Offline sparse head execution simulation from existing DDIM logs."
+    )
+    parser.add_argument(
+        "--root_dir",
+        type=str,
+        default=".",
+        help="Repository root or directory under which case outputs are stored.",
+    )
+    parser.add_argument(
+        "--output_dir",
+        type=str,
+        default=None,
+        help="Directory to write sparse-head simulation outputs. Defaults to <root_dir>/sparse_head_simulation.",
+    )
+    parser.add_argument(
+        "--target_threshold",
+        type=float,
+        default=0.95,
+        help="Target cosine threshold used by experiment 1.2.",
+    )
+    return parser.parse_args()
+
+
+def discover_stepwise_logs(root_dir: Path) -> list[Path]:
+    return sorted(root_dir.glob("unitree_*/case*/output/inference/stepwise_log.csv"))
+
+
+def load_stepwise_table(stepwise_paths: list[Path]) -> pd.DataFrame:
+    frames = []
+    for path in stepwise_paths:
+        frame = pd.read_csv(path)
+        frame["dataset"] = path.parts[-5]
+        frame["case"] = path.parts[-4]
+        frames.append(frame)
+    if not frames:
+        raise FileNotFoundError("No stepwise_log.csv files found.")
+    stepwise_df = pd.concat(frames, ignore_index=True)
+    for column in [
+            "step",
+            "step_time_s",
+            "latent_delta",
+            "action_delta",
+            "state_delta",
+            "action_cosine_vs_full50",
+            "state_cosine_vs_full50",
+            "latent_l2_vs_full50",
+    ]:
+        stepwise_df[column] = pd.to_numeric(stepwise_df[column], errors="coerce")
+    return stepwise_df
+
+
+def simulate_schemes(stepwise_df: pd.DataFrame,
+                     schemes: dict[str, list[int]]) -> pd.DataFrame:
+    rows = []
+    group_columns = [
+        "dataset",
+        "case",
+        "sample_id",
+        "scene",
+        "pass_type",
+        "round_id",
+    ]
+    grouped = stepwise_df.groupby(group_columns)
+    for keys, group in grouped:
+        group = group.sort_values("step").reset_index(drop=True)
+        action_curve = dict(zip(group["step"] - 1, group["action_cosine_vs_full50"]))
+        state_curve = dict(zip(group["step"] - 1, group["state_cosine_vs_full50"]))
+
+        for scheme_name, checkpoints in schemes.items():
+            normalized_checkpoints = sorted(
+                checkpoint for checkpoint in checkpoints if 0 <= checkpoint <= 49)
+            if not normalized_checkpoints:
+                continue
+            last_checkpoint = normalized_checkpoints[-1]
+            rows.append({
+                "dataset": keys[0],
+                "case": keys[1],
+                "sample_id": keys[2],
+                "scene": keys[3],
+                "pass_type": keys[4],
+                "round_id": keys[5],
+                "scheme": scheme_name,
+                "head_exec_steps_zero_based": json.dumps(normalized_checkpoints),
+                "head_exec_count": len(normalized_checkpoints),
+                "head_compute_saving_ratio": 1.0 - len(normalized_checkpoints) / 50.0,
+                "final_checkpoint_zero_based": last_checkpoint,
+                "final_checkpoint_one_based": last_checkpoint + 1,
+                "final_action_cosine_vs_dense": action_curve[last_checkpoint],
+                "final_state_cosine_vs_dense": state_curve[last_checkpoint],
+            })
+    return pd.DataFrame(rows)
+
+
+def summarize_scheme_results(simulation_df: pd.DataFrame) -> tuple[pd.DataFrame, pd.DataFrame]:
+    metric_columns = [
+        "final_action_cosine_vs_dense",
+        "final_state_cosine_vs_dense",
+        "head_exec_count",
+        "head_compute_saving_ratio",
+    ]
+    overall = simulation_df.groupby(["scheme", "pass_type"])[metric_columns].agg(
+        ["mean", "median", "std"]).reset_index()
+    per_case = simulation_df.groupby(
+        ["dataset", "case", "scheme", "pass_type"])[metric_columns].mean().reset_index()
+    return overall, per_case
+
+
+def compute_min_steps_needed(stepwise_df: pd.DataFrame,
+                             threshold: float) -> pd.DataFrame:
+    rows = []
+    group_columns = [
+        "dataset",
+        "case",
+        "sample_id",
+        "scene",
+        "pass_type",
+        "round_id",
+    ]
+    grouped = stepwise_df.groupby(group_columns)
+    for keys, group in grouped:
+        group = group.sort_values("step").reset_index(drop=True)
+        action_hits = group[group["action_cosine_vs_full50"] >= threshold]
+        state_hits = group[group["state_cosine_vs_full50"] >= threshold]
+
+        action_step = np.nan if action_hits.empty else float(action_hits.iloc[0]["step"] - 1)
+        state_step = np.nan if state_hits.empty else float(state_hits.iloc[0]["step"] - 1)
+        rows.append({
+            "dataset": keys[0],
+            "case": keys[1],
+            "sample_id": keys[2],
+            "scene": keys[3],
+            "pass_type": keys[4],
+            "round_id": keys[5],
+            "target_threshold": threshold,
+            "action_min_head_steps_needed": 1.0 if pd.notna(action_step) else np.nan,
+            "state_min_head_steps_needed": 1.0 if pd.notna(state_step) else np.nan,
+            "action_earliest_checkpoint_zero_based": action_step,
+            "state_earliest_checkpoint_zero_based": state_step,
+            "action_max_head_compute_saving_ratio": 0.98 if pd.notna(action_step) else np.nan,
+            "state_max_head_compute_saving_ratio": 0.98 if pd.notna(state_step) else np.nan,
+        })
+    return pd.DataFrame(rows)
+
+
+def compare_tail_vs_uniform(simulation_df: pd.DataFrame) -> pd.DataFrame:
+    comparison_specs = [
+        ("sparse_8", "tail_heavy_8", "budget_8_uniform_vs_tail"),
+        ("sparse_4", "tail_only_4", "budget_4_uniform_vs_tail"),
+        ("sparse_10", "tail_heavy_10", "budget_10_uniform_vs_tail"),
+    ]
+    base_columns = [
+        "dataset",
+        "case",
+        "sample_id",
+        "scene",
+        "pass_type",
+        "round_id",
+    ]
+    rows = []
+    for uniform_scheme, tail_scheme, comparison_name in comparison_specs:
+        uniform_df = simulation_df[simulation_df["scheme"] == uniform_scheme].copy()
+        tail_df = simulation_df[simulation_df["scheme"] == tail_scheme].copy()
+        merged = uniform_df.merge(
+            tail_df,
+            on=base_columns,
+            suffixes=("_uniform", "_tail"),
+            how="inner",
+        )
+        for _, row in merged.iterrows():
+            rows.append({
+                "comparison": comparison_name,
+                "dataset": row["dataset"],
+                "case": row["case"],
+                "sample_id": row["sample_id"],
+                "scene": row["scene"],
+                "pass_type": row["pass_type"],
+                "round_id": row["round_id"],
+                "uniform_scheme": uniform_scheme,
+                "tail_scheme": tail_scheme,
+                "uniform_head_exec_count": row["head_exec_count_uniform"],
+                "tail_head_exec_count": row["head_exec_count_tail"],
+                "uniform_action_cosine_vs_dense": row["final_action_cosine_vs_dense_uniform"],
+                "tail_action_cosine_vs_dense": row["final_action_cosine_vs_dense_tail"],
+                "uniform_state_cosine_vs_dense": row["final_state_cosine_vs_dense_uniform"],
+                "tail_state_cosine_vs_dense": row["final_state_cosine_vs_dense_tail"],
+                "tail_minus_uniform_action_cosine":
+                row["final_action_cosine_vs_dense_tail"] -
+                row["final_action_cosine_vs_dense_uniform"],
+                "tail_minus_uniform_state_cosine":
+                row["final_state_cosine_vs_dense_tail"] -
+                row["final_state_cosine_vs_dense_uniform"],
+                "tail_better_action": row["final_action_cosine_vs_dense_tail"] >
+                row["final_action_cosine_vs_dense_uniform"],
+                "tail_better_state": row["final_state_cosine_vs_dense_tail"] >
+                row["final_state_cosine_vs_dense_uniform"],
+            })
+    return pd.DataFrame(rows)
+
+
+def build_summary_payload(simulation_df: pd.DataFrame,
+                          min_steps_df: pd.DataFrame,
+                          tail_compare_df: pd.DataFrame,
+                          target_threshold: float) -> dict:
+    payload: dict[str, dict] = {
+        "config": {
+            "target_threshold": target_threshold,
+            "schemes": DEFAULT_SCHEMES,
+        },
+        "experiment_1_1": {},
+        "experiment_1_2": {},
+        "experiment_1_3": {},
+    }
+
+    for scheme, group in simulation_df.groupby("scheme"):
+        payload["experiment_1_1"][scheme] = {
+            "num_rows": int(len(group)),
+            "head_exec_count": float(group["head_exec_count"].iloc[0]),
+            "head_compute_saving_ratio": float(
+                group["head_compute_saving_ratio"].iloc[0]),
+            "final_action_cosine_vs_dense_mean": float(
+                group["final_action_cosine_vs_dense"].mean()),
+            "final_action_cosine_vs_dense_median": float(
+                group["final_action_cosine_vs_dense"].median()),
+            "final_state_cosine_vs_dense_mean": float(
+                group["final_state_cosine_vs_dense"].mean()),
+            "final_state_cosine_vs_dense_median": float(
+                group["final_state_cosine_vs_dense"].median()),
+        }
+
+    payload["experiment_1_2"] = {
+        "action_earliest_checkpoint_zero_based": {
+            "mean": float(
+                pd.to_numeric(
+                    min_steps_df["action_earliest_checkpoint_zero_based"],
+                    errors="coerce").dropna().mean()),
+            "median": float(
+                pd.to_numeric(
+                    min_steps_df["action_earliest_checkpoint_zero_based"],
+                    errors="coerce").dropna().median()),
+        },
+        "state_earliest_checkpoint_zero_based": {
+            "mean": float(
+                pd.to_numeric(
+                    min_steps_df["state_earliest_checkpoint_zero_based"],
+                    errors="coerce").dropna().mean()),
+            "median": float(
+                pd.to_numeric(
+                    min_steps_df["state_earliest_checkpoint_zero_based"],
+                    errors="coerce").dropna().median()),
+        },
+        "min_head_steps_needed_action_unique": sorted(
+            pd.to_numeric(min_steps_df["action_min_head_steps_needed"],
+                          errors="coerce").dropna().unique().tolist()),
+        "min_head_steps_needed_state_unique": sorted(
+            pd.to_numeric(min_steps_df["state_min_head_steps_needed"],
+                          errors="coerce").dropna().unique().tolist()),
+    }
+
+    for comparison, group in tail_compare_df.groupby("comparison"):
+        payload["experiment_1_3"][comparison] = {
+            "num_rows": int(len(group)),
+            "tail_better_action_share": float(group["tail_better_action"].mean()),
+            "tail_better_state_share": float(group["tail_better_state"].mean()),
+            "tail_minus_uniform_action_cosine_mean": float(
+                group["tail_minus_uniform_action_cosine"].mean()),
+            "tail_minus_uniform_state_cosine_mean": float(
+                group["tail_minus_uniform_state_cosine"].mean()),
+        }
+    return payload
+
+
+def write_markdown_report(path: Path, payload: dict) -> None:
+    lines = [
+        "# Sparse Head Execution Simulation",
+        "",
+        "This report uses zero-order hold over logged stepwise action/state outputs.",
+        "For a sparse scheme, the final output at step 49 is approximated by the most recent checkpoint output.",
+        "",
+        "## Experiment 1.1",
+        "",
+    ]
+
+    for scheme, stats in payload["experiment_1_1"].items():
+        lines.extend([
+            f"### {scheme}",
+            "",
+            f"- Head exec count: {stats['head_exec_count']:.0f}",
+            f"- Head compute saving ratio: {stats['head_compute_saving_ratio']:.4f}",
+            f"- Final action cosine vs dense: mean={stats['final_action_cosine_vs_dense_mean']:.4f}, median={stats['final_action_cosine_vs_dense_median']:.4f}",
+            f"- Final state cosine vs dense: mean={stats['final_state_cosine_vs_dense_mean']:.4f}, median={stats['final_state_cosine_vs_dense_median']:.4f}",
+            "",
+        ])
+
+    lines.extend([
+        "## Experiment 1.2",
+        "",
+        f"- Target threshold: {payload['config']['target_threshold']}",
+        f"- Action earliest checkpoint mean: {payload['experiment_1_2']['action_earliest_checkpoint_zero_based']['mean']:.4f}",
+        f"- State earliest checkpoint mean: {payload['experiment_1_2']['state_earliest_checkpoint_zero_based']['mean']:.4f}",
+        f"- Unique min head steps needed for action: {payload['experiment_1_2']['min_head_steps_needed_action_unique']}",
+        f"- Unique min head steps needed for state: {payload['experiment_1_2']['min_head_steps_needed_state_unique']}",
+        "",
+        "## Experiment 1.3",
+        "",
+    ])
+
+    for comparison, stats in payload["experiment_1_3"].items():
+        lines.extend([
+            f"### {comparison}",
+            "",
+            f"- Tail better action share: {stats['tail_better_action_share']:.4f}",
+            f"- Tail better state share: {stats['tail_better_state_share']:.4f}",
+            f"- Mean tail-minus-uniform action cosine: {stats['tail_minus_uniform_action_cosine_mean']:.4f}",
+            f"- Mean tail-minus-uniform state cosine: {stats['tail_minus_uniform_state_cosine_mean']:.4f}",
+            "",
+        ])
+
+    with open(path, "w", encoding="utf-8") as file:
+        file.write("\n".join(lines) + "\n")
+
+
+def main() -> None:
+    args = parse_args()
+    root_dir = Path(args.root_dir).resolve()
+    output_dir = Path(args.output_dir).resolve(
+    ) if args.output_dir else root_dir / "sparse_head_simulation"
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+    stepwise_paths = discover_stepwise_logs(root_dir)
+    stepwise_df = load_stepwise_table(stepwise_paths)
+
+    simulation_df = simulate_schemes(stepwise_df, DEFAULT_SCHEMES)
+    scheme_overall_df, scheme_per_case_df = summarize_scheme_results(simulation_df)
+    min_steps_df = compute_min_steps_needed(stepwise_df, args.target_threshold)
+    tail_compare_df = compare_tail_vs_uniform(simulation_df)
+    summary_payload = build_summary_payload(simulation_df, min_steps_df,
+                                            tail_compare_df,
+                                            args.target_threshold)
+
+    simulation_df.to_csv(output_dir / "scheme_simulation_per_round.csv",
+                         index=False)
+    scheme_overall_df.to_csv(output_dir / "scheme_simulation_overall.csv",
+                             index=False)
+    scheme_per_case_df.to_csv(output_dir / "scheme_simulation_per_case.csv",
+                              index=False)
+    min_steps_df.to_csv(output_dir / "min_head_steps_needed.csv", index=False)
+    tail_compare_df.to_csv(output_dir / "tail_vs_uniform.csv", index=False)
+    with open(output_dir / "summary.json", "w", encoding="utf-8") as file:
+        json.dump(summary_payload, file, indent=2, ensure_ascii=False)
+    write_markdown_report(output_dir / "report.md", summary_payload)
+
+    print(f"Sparse head simulation written to: {output_dir}")
+
+
+if __name__ == "__main__":
+    main()

scripts/export_trt.py

@@ -0,0 +1,209 @@
+"""Export video UNet backbone to ONNX, then convert to TensorRT engine.
+
+Usage:
+    python scripts/export_trt.py \
+        --ckpt ckpts/unifolm_wma_dual.ckpt.prepared.pt \
+        --config configs/inference/world_model_interaction.yaml \
+        --out_dir trt_engines
+
+    python scripts/export_trt.py \
+        --ckpt ckpts/unifolm_wma_dual.ckpt.prepared.pt \
+        --config configs/inference/world_model_interaction.yaml \
+        --engine_path trt_engines/video_backbone_multigpu.engine \
+        --onnx_path trt_engines/video_backbone_multigpu.onnx
+"""
+
+import os
+import sys
+import argparse
+import json
+
+import torch
+import tensorrt as trt
+from omegaconf import OmegaConf
+
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'src'))
+from unifolm_wma.utils.utils import instantiate_from_config
+from unifolm_wma.trt_utils import export_backbone_onnx
+
+
+class TacticRecorder(trt.IAlgorithmSelector):
+    """Pass 1: record all candidate tactics and the auto-selected winner."""
+
+    def __init__(self):
+        super().__init__()
+        self.records = {}  # layer_name -> {candidates: [...], selected: ...}
+
+    def select_algorithms(self, ctx, choices):
+        name = ctx.name
+        # Collect input/output shapes
+        inputs = []
+        for j in range(ctx.num_inputs):
+            try:
+                inputs.append([int(d) for d in ctx.get_shape(j)])
+            except Exception:
+                inputs.append(None)
+        outputs = []
+        for j in range(ctx.num_outputs):
+            try:
+                outputs.append([int(d) for d in ctx.get_shape(ctx.num_inputs + j)])
+            except Exception:
+                outputs.append(None)
+        self.records[name] = {
+            "input_shapes": inputs,
+            "output_shapes": outputs,
+            "candidates": [],
+            "selected": None,
+        }
+        for i, c in enumerate(choices):
+            v = c.algorithm_variant
+            self.records[name]["candidates"].append({
+                "index": i,
+                "implementation": v.implementation,
+                "tactic": v.tactic,
+                "timing_msec": c.timing_msec,
+                "workspace_size": c.workspace_size,
+            })
+        # return all indices -> let TRT auto-pick the fastest
+        return list(range(len(choices)))
+
+    def report_algorithms(self, ctx, choices):
+        # Both ctx and choices are lists in report_algorithms
+        for c, alg in zip(ctx, choices):
+            name = c.name
+            if name in self.records:
+                v = alg.algorithm_variant
+                self.records[name]["selected"] = {
+                    "implementation": v.implementation,
+                    "tactic": v.tactic,
+                    "timing_msec": alg.timing_msec,
+                    "workspace_size": alg.workspace_size,
+                }
+
+    def save(self, path):
+        with open(path, "w") as f:
+            json.dump(self.records, f, indent=2)
+        print(f">>> Tactic info saved to {path} ({len(self.records)} layers)")
+
+
+class TacticForcer(trt.IAlgorithmSelector):
+    """Pass 2: force user-specified tactics from a JSON file."""
+
+    def __init__(self, path):
+        super().__init__()
+        with open(path) as f:
+            self.overrides = json.load(f)
+        n = sum(1 for v in self.overrides.values() if v.get("force"))
+        print(f">>> Loaded tactic overrides: {n} layers with 'force' set")
+
+    def select_algorithms(self, ctx, choices):
+        name = ctx.name
+        override = self.overrides.get(name)
+        if override and override.get("force"):
+            target_impl = override["force"]["implementation"]
+            target_tactic = override["force"]["tactic"]
+            for i, c in enumerate(choices):
+                v = c.algorithm_variant
+                if v.implementation == target_impl and v.tactic == target_tactic:
+                    return [i]
+            print(f"    WARN: forced tactic not found for {name}, using auto")
+        return list(range(len(choices)))
+
+    def report_algorithms(self, ctx, choices):
+        pass
+
+
+def load_model(config_path, ckpt_path, device):
+    if ckpt_path.endswith('.prepared.pt'):
+        model = torch.load(ckpt_path, map_location='cpu')
+    else:
+        config = OmegaConf.load(config_path)
+        model = instantiate_from_config(config.model)
+        state_dict = torch.load(ckpt_path, map_location='cpu')
+        if 'state_dict' in state_dict:
+            state_dict = state_dict['state_dict']
+        model.load_state_dict(state_dict, strict=False)
+    model.eval().to(device)
+    return model
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--ckpt', required=True)
+    parser.add_argument('--config', default='configs/inference/world_model_interaction.yaml')
+    parser.add_argument('--out_dir', default='trt_engines')
+    parser.add_argument('--gpu_id',
+                        type=int,
+                        default=0,
+                        help='CUDA device id used for ONNX export and TRT build.')
+    parser.add_argument('--onnx_path',
+                        default=None,
+                        help='Optional explicit ONNX output path. Overrides --out_dir default name.')
+    parser.add_argument('--engine_path',
+                        default=None,
+                        help='Optional explicit TensorRT engine output path. Overrides --out_dir default name.')
+    parser.add_argument('--context_len', type=int, default=95)
+    parser.add_argument('--fp16', action='store_true', default=True)
+    parser.add_argument('--dump-tactics', default=None, help='Pass 1: dump tactic info to JSON')
+    parser.add_argument('--load-tactics', default=None, help='Pass 2: force tactics from JSON')
+    args = parser.parse_args()
+    device = torch.device('cuda', args.gpu_id)
+    torch.cuda.set_device(device)
+
+    onnx_path = args.onnx_path or os.path.join(args.out_dir,
+                                               'video_backbone.onnx')
+    engine_path = args.engine_path or os.path.join(args.out_dir,
+                                                   'video_backbone.engine')
+    os.makedirs(os.path.dirname(os.path.abspath(onnx_path)), exist_ok=True)
+    os.makedirs(os.path.dirname(os.path.abspath(engine_path)), exist_ok=True)
+
+    if os.path.exists(onnx_path):
+        print(f">>> ONNX already exists at {onnx_path}, skipping export.")
+        n_outputs = 10
+    else:
+        print(">>> Loading model ...")
+        model = load_model(args.config, args.ckpt, device)
+        print(">>> Exporting ONNX ...")
+        with torch.no_grad():
+            n_outputs = export_backbone_onnx(model, onnx_path, context_len=args.context_len)
+        del model
+        torch.cuda.empty_cache()
+
+    print(">>> Converting ONNX -> TensorRT engine ...")
+    logger = trt.Logger(trt.Logger.WARNING)
+    builder = trt.Builder(logger)
+    network = builder.create_network(1 << int(trt.NetworkDefinitionCreationFlag.EXPLICIT_BATCH))
+    parser = trt.OnnxParser(network, logger)
+
+    if not parser.parse_from_file(os.path.abspath(onnx_path)):
+        for i in range(parser.num_errors):
+            print(f"    ONNX parse error: {parser.get_error(i)}")
+        raise RuntimeError("ONNX parsing failed")
+
+    config = builder.create_builder_config()
+    config.set_memory_pool_limit(trt.MemoryPoolType.WORKSPACE, 16 << 30)
+    if args.fp16:
+        config.set_flag(trt.BuilderFlag.FP16)
+
+    # Tactic selection
+    recorder = None
+    if args.dump_tactics:
+        recorder = TacticRecorder()
+        config.algorithm_selector = recorder
+    elif args.load_tactics:
+        config.algorithm_selector = TacticForcer(args.load_tactics)
+
+    engine_bytes = builder.build_serialized_network(network, config)
+
+    if recorder and args.dump_tactics:
+        recorder.save(args.dump_tactics)
+
+    with open(engine_path, 'wb') as f:
+        f.write(engine_bytes)
+
+    print(f"\n>>> Done! Engine saved to {engine_path}")
+    print(f"    Outputs: 1 y + {n_outputs - 1} hs_a tensors")
+
+
+if __name__ == '__main__':
+    main()

scripts/run_all_world_model_cases.sh

@@ -0,0 +1,60 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+
+repo_root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${repo_root}"
+
+run_analysis="${RUN_ANALYSIS:-1}"
+case_filter="${CASE_FILTER:-}"
+python_bin="${PYTHON_BIN:-}"
+
+if [[ -z "${python_bin}" ]]; then
+    if [[ -n "${CONDA_PREFIX:-}" ]] && [[ -x "${CONDA_PREFIX}/bin/python" ]]; then
+        python_bin="${CONDA_PREFIX}/bin/python"
+    elif [[ -n "${VIRTUAL_ENV:-}" ]] && [[ -x "${VIRTUAL_ENV}/bin/python" ]]; then
+        python_bin="${VIRTUAL_ENV}/bin/python"
+    else
+        python_bin="$(command -v python || true)"
+    fi
+fi
+
+if [[ -z "${python_bin}" ]] || [[ ! -x "${python_bin}" ]]; then
+    echo "Unable to resolve a usable Python interpreter. Set PYTHON_BIN explicitly." >&2
+    exit 1
+fi
+
+case_scripts=(
+    "unitree_z1_stackbox/case1/run_world_model_interaction.sh"
+    "unitree_z1_stackbox/case2/run_world_model_interaction.sh"
+    "unitree_z1_stackbox/case3/run_world_model_interaction.sh"
+    "unitree_z1_stackbox/case4/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_stackbox/case1/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_stackbox/case2/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_stackbox/case3/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_stackbox/case4/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_stackbox_v2/case1/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_stackbox_v2/case2/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_stackbox_v2/case3/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_stackbox_v2/case4/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_cleanup_pencils/case1/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_cleanup_pencils/case2/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_cleanup_pencils/case3/run_world_model_interaction.sh"
+    "unitree_z1_dual_arm_cleanup_pencils/case4/run_world_model_interaction.sh"
+    "unitree_g1_pack_camera/case1/run_world_model_interaction.sh"
+    "unitree_g1_pack_camera/case2/run_world_model_interaction.sh"
+    "unitree_g1_pack_camera/case3/run_world_model_interaction.sh"
+    "unitree_g1_pack_camera/case4/run_world_model_interaction.sh"
+)
+
+for case_script in "${case_scripts[@]}"; do
+    if [[ -n "${case_filter}" ]] && [[ "${case_script}" != *"${case_filter}"* ]]; then
+        continue
+    fi
+
+    echo "============================================================"
+    echo "Running ${case_script}"
+    echo "============================================================"
+    echo "Using Python: ${python_bin}"
+    RUN_ANALYSIS="${run_analysis}" PYTHON_BIN="${python_bin}" bash "${repo_root}/${case_script}"
+done

scripts/run_backbone_reuse_worst_cases.sh

@@ -0,0 +1,20 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+
+repo_root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${repo_root}"
+
+case_scripts=(
+    "unitree_z1_dual_arm_stackbox/case1/run_world_model_backbone_reuse_output546.sh"
+    "unitree_z1_dual_arm_stackbox/case3/run_world_model_backbone_reuse_output546.sh"
+    "unitree_z1_dual_arm_stackbox_v2/case2/run_world_model_backbone_reuse_output546.sh"
+    "unitree_z1_dual_arm_stackbox_v2/case3/run_world_model_backbone_reuse_output546.sh"
+)
+
+for case_script in "${case_scripts[@]}"; do
+    echo "============================================================"
+    echo "Running ${case_script}"
+    echo "============================================================"
+    bash "${repo_root}/${case_script}"
+done

scripts/run_case3_backbone_profiles.sh

@@ -0,0 +1,53 @@
+#!/usr/bin/env bash
+
+set -uo pipefail
+
+repo_root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${repo_root}"
+
+gpu_devices="${CUDA_VISIBLE_DEVICES:-1}"
+python_bin="${PYTHON_BIN:-}"
+
+if [[ -z "${python_bin}" ]]; then
+    if [[ -n "${CONDA_PREFIX:-}" ]] && [[ -x "${CONDA_PREFIX}/bin/python" ]]; then
+        python_bin="${CONDA_PREFIX}/bin/python"
+    elif [[ -n "${VIRTUAL_ENV:-}" ]] && [[ -x "${VIRTUAL_ENV}/bin/python" ]]; then
+        python_bin="${VIRTUAL_ENV}/bin/python"
+    else
+        python_bin="$(command -v python || true)"
+    fi
+fi
+
+if [[ -z "${python_bin}" ]] || [[ ! -x "${python_bin}" ]]; then
+    echo "Unable to resolve a usable Python interpreter. Set PYTHON_BIN explicitly." >&2
+    exit 1
+fi
+
+case_scripts=(
+    "unitree_z1_dual_arm_stackbox/case3/run_world_model_backbone_profile.sh"
+    "unitree_z1_dual_arm_stackbox_v2/case3/run_world_model_backbone_profile.sh"
+    "unitree_z1_dual_arm_cleanup_pencils/case3/run_world_model_backbone_profile.sh"
+    "unitree_g1_pack_camera/case3/run_world_model_backbone_profile.sh"
+)
+
+failed_cases=()
+
+for case_script in "${case_scripts[@]}"; do
+    echo "============================================================"
+    echo "Running ${case_script} on GPU ${gpu_devices}"
+    echo "============================================================"
+    echo "Using Python: ${python_bin}"
+    if ! CUDA_VISIBLE_DEVICES="${gpu_devices}" PYTHON_BIN="${python_bin}" bash "${repo_root}/${case_script}"; then
+        echo "FAILED: ${case_script}" >&2
+        failed_cases+=("${case_script}")
+    fi
+done
+
+if [[ ${#failed_cases[@]} -gt 0 ]]; then
+    echo "============================================================" >&2
+    echo "Backbone profiling finished with failures:" >&2
+    for case_script in "${failed_cases[@]}"; do
+        echo "  - ${case_script}" >&2
+    done
+    exit 1
+fi

scripts/run_world_model_backbone_profile_case.sh

@@ -0,0 +1,86 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+
+if [[ $# -ne 4 ]]; then
+    echo "Usage: $0 <res_dir> <dataset> <frame_stride> <n_iter>" >&2
+    exit 1
+fi
+
+res_dir="$1"
+dataset="$2"
+frame_stride="$3"
+n_iter="$4"
+
+repo_root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${repo_root}"
+
+gpu_devices="${CUDA_VISIBLE_DEVICES:-1}"
+python_bin="${PYTHON_BIN:-}"
+if [[ -z "${python_bin}" ]]; then
+    if [[ -n "${CONDA_PREFIX:-}" ]] && [[ -x "${CONDA_PREFIX}/bin/python" ]]; then
+        python_bin="${CONDA_PREFIX}/bin/python"
+    elif [[ -n "${VIRTUAL_ENV:-}" ]] && [[ -x "${VIRTUAL_ENV}/bin/python" ]]; then
+        python_bin="${VIRTUAL_ENV}/bin/python"
+    else
+        python_bin="$(command -v python || true)"
+    fi
+fi
+if [[ -z "${python_bin}" ]] || [[ ! -x "${python_bin}" ]]; then
+    echo "Unable to resolve a usable Python interpreter. Set PYTHON_BIN explicitly." >&2
+    exit 1
+fi
+save_dir="${res_dir}/backbone_profile_dense"
+log_path="${save_dir}/run.log"
+
+mkdir -p "${save_dir}"
+
+interaction_cmd=(
+    "${python_bin}" scripts/evaluation/world_model_interaction.py
+    --seed 123
+    --ckpt_path ckpts/unifolm_wma_dual.ckpt
+    --config configs/inference/world_model_interaction.yaml
+    --savedir "${save_dir}"
+    --bs 1
+    --height 320
+    --width 512
+    --unconditional_guidance_scale 1.0
+    --ddim_steps 50
+    --ddim_eta 1.0
+    --prompt_dir "${res_dir}/world_model_interaction_prompts"
+    --dataset "${dataset}"
+    --video_length 16
+    --frame_stride "${frame_stride}"
+    --n_action_steps 16
+    --exe_steps 16
+    --n_iter "${n_iter}"
+    --timestep_spacing uniform_trailing
+    --guidance_rescale 0.7
+    --perframe_ae
+    --analysis_log_metrics
+    --analysis_reference_steps 50
+    --analysis_profile_backbone_blocks
+)
+
+metrics_analysis_cmd=(
+    "${python_bin}" scripts/evaluation/analyze_metrics.py
+    --input_dir "${save_dir}/inference"
+    --output_dir "${save_dir}/inference/analysis"
+)
+
+backbone_analysis_cmd=(
+    "${python_bin}" scripts/evaluation/analyze_backbone_blocks.py
+    --input_csv "${save_dir}/inference/backbone_block_log.csv"
+    --output_dir "${save_dir}/inference/backbone_analysis"
+)
+
+{
+    echo "============================================================"
+    echo "Running dense backbone profiling for ${res_dir} on GPU ${gpu_devices}"
+    echo "============================================================"
+    echo "Using Python: ${python_bin}"
+    time env CUDA_VISIBLE_DEVICES="${gpu_devices}" "${interaction_cmd[@]}"
+
+    "${metrics_analysis_cmd[@]}"
+    "${backbone_analysis_cmd[@]}"
+} 2>&1 | tee "${log_path}"

scripts/run_world_model_case_schemes.sh

@@ -0,0 +1,106 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+
+if [[ $# -ne 4 ]]; then
+    echo "Usage: $0 <res_dir> <dataset> <frame_stride> <n_iter>" >&2
+    exit 1
+fi
+
+res_dir="$1"
+dataset="$2"
+frame_stride="$3"
+n_iter="$4"
+
+repo_root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${repo_root}"
+
+run_analysis="${RUN_ANALYSIS:-1}"
+scheme_filter="${SCHEME_FILTER:-}"
+gpu_devices="${CUDA_VISIBLE_DEVICES:-0}"
+python_bin="${PYTHON_BIN:-}"
+if [[ -z "${python_bin}" ]]; then
+    if [[ -n "${CONDA_PREFIX:-}" ]] && [[ -x "${CONDA_PREFIX}/bin/python" ]]; then
+        python_bin="${CONDA_PREFIX}/bin/python"
+    elif [[ -n "${VIRTUAL_ENV:-}" ]] && [[ -x "${VIRTUAL_ENV}/bin/python" ]]; then
+        python_bin="${VIRTUAL_ENV}/bin/python"
+    else
+        python_bin="$(command -v python || true)"
+    fi
+fi
+if [[ -z "${python_bin}" ]] || [[ ! -x "${python_bin}" ]]; then
+    echo "Unable to resolve a usable Python interpreter. Set PYTHON_BIN explicitly." >&2
+    exit 1
+fi
+head_log_steps=(40 43 46 47 48 49)
+scheme_names=(sparse_8)
+
+get_head_schedule() {
+    case "$1" in
+        sparse_8)
+            echo "0 7 14 21 28 35 42 49"
+            ;;
+        sparse_4)
+            echo "0 16 32 49"
+            ;;
+        tail_only)
+            echo "40 43 46 49"
+            ;;
+        tail_heavy)
+            echo "0 32 40 44 47 49"
+            ;;
+        *)
+            echo "Unknown scheme: $1" >&2
+            exit 1
+            ;;
+    esac
+}
+
+mkdir -p "${res_dir}/output_logs"
+
+for scheme_name in "${scheme_names[@]}"; do
+    if [[ -n "${scheme_filter}" ]] && [[ "${scheme_name}" != *"${scheme_filter}"* ]]; then
+        continue
+    fi
+
+    read -r -a head_schedule <<< "$(get_head_schedule "${scheme_name}")"
+    scheme_savedir="${res_dir}/output/${scheme_name}"
+    scheme_log_path="${res_dir}/output_logs/${scheme_name}.log"
+
+    {
+        echo "============================================================"
+        echo "Running ${res_dir} with ${scheme_name}"
+        echo "============================================================"
+        echo "Using Python: ${python_bin}"
+        time CUDA_VISIBLE_DEVICES="${gpu_devices}" "${python_bin}" scripts/evaluation/world_model_interaction.py \
+            --seed 123 \
+            --ckpt_path ckpts/unifolm_wma_dual.ckpt \
+            --config configs/inference/world_model_interaction.yaml \
+            --savedir "${scheme_savedir}" \
+            --bs 1 --height 320 --width 512 \
+            --unconditional_guidance_scale 1.0 \
+            --ddim_steps 50 \
+            --ddim_eta 1.0 \
+            --prompt_dir "${res_dir}/world_model_interaction_prompts" \
+            --dataset "${dataset}" \
+            --video_length 16 \
+            --frame_stride "${frame_stride}" \
+            --n_action_steps 16 \
+            --exe_steps 16 \
+            --n_iter "${n_iter}" \
+            --timestep_spacing uniform_trailing \
+            --guidance_rescale 0.7 \
+            --perframe_ae \
+            --analysis_log_metrics \
+            --analysis_reference_steps 50 \
+            --head_schedule_steps "${head_schedule[@]}" \
+            --head_skip_mode reuse_prediction \
+            --head_log_steps "${head_log_steps[@]}"
+
+        if [[ "${run_analysis}" == "1" ]]; then
+            "${python_bin}" scripts/evaluation/analyze_metrics.py \
+                --input_dir "${scheme_savedir}/inference" \
+                --output_dir "${scheme_savedir}/inference/analysis"
+        fi
+    } 2>&1 | tee "${scheme_log_path}"
+done

scripts/run_world_model_interaction.sh

@@ -3,6 +3,8 @@ ckpt=/path/to/model/checkpoint
 config=configs/inference/world_model_interaction.yaml
 seed=123
 res_dir="/path/to/result/directory"
+analysis_reference_steps=50
+# analysis_psnr_path="/path/to/psnr_full50.csv"
 
 datasets=(
     "unitree_z1_stackbox"
@@ -19,6 +21,13 @@ for i in "${!datasets[@]}"; do
     dataset=${datasets[$i]}
     n_iter=${n_iters[$i]}
     fs=${fses[$i]}
+    analysis_args=(
+        --analysis_log_metrics
+        --analysis_reference_steps ${analysis_reference_steps}
+    )
+    if [ -n "${analysis_psnr_path:-}" ]; then
+        analysis_args+=(--analysis_psnr_path "${analysis_psnr_path}")
+    fi
 
     CUDA_VISIBLE_DEVICES=0 python3 scripts/evaluation/world_model_interaction.py \
     --seed ${seed} \
@@ -38,5 +47,6 @@ for i in "${!datasets[@]}"; do
     --n_iter ${n_iter} \
     --timestep_spacing 'uniform_trailing' \
     --guidance_rescale 0.7 \
-    --perframe_ae
+    --perframe_ae \
+    "${analysis_args[@]}"
 done

sparse_head_simulation/min_head_steps_needed.csv

@@ -0,0 +1,393 @@
+dataset,case,sample_id,scene,pass_type,round_id,target_threshold,action_min_head_steps_needed,state_min_head_steps_needed,action_earliest_checkpoint_zero_based,state_earliest_checkpoint_zero_based,action_max_head_compute_saving_ratio,state_max_head_compute_saving_ratio
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,0,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,1,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,3,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,4,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,5,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,6,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,7,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,8,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,9,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,10,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,0,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,3,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,4,0.95,1.0,1.0,46.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,5,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,6,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,7,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,8,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,9,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_g1_pack_camera,case1,unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,10,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,0,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,1,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,3,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,4,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,5,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,9,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,10,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,1,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,3,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,4,0.95,1.0,1.0,46.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,5,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,8,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,9,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case2,unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,10,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,0,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,1,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,3,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,4,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,7,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,8,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,9,0.95,1.0,1.0,46.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,10,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,0,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,3,0.95,1.0,1.0,46.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,4,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,5,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,7,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,8,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,9,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case3,unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,10,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,0,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,1,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,3,0.95,1.0,1.0,46.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,4,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,7,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,8,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,9,0.95,1.0,1.0,46.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,policy,10,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,3,0.95,1.0,1.0,46.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,4,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,5,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,6,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,7,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,8,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,9,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_g1_pack_camera,case4,unitree_g1_pack_camera-vid200-fs6,unitree_g1_pack_camera,world_model,10,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,0,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,4,0.95,1.0,1.0,44.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,5,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,6,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,7,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,0,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,4,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,5,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,6,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case1,unitree_z1_dual_arm_cleanup_pencils-vid0-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,0,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,3,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,4,0.95,1.0,1.0,44.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,5,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,6,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,0,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,3,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,4,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,5,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,6,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case2,unitree_z1_dual_arm_cleanup_pencils-vid50-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,0,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,3,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,4,0.95,1.0,1.0,44.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,6,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,0,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,3,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,4,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,5,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,6,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case3,unitree_z1_dual_arm_cleanup_pencils-vid100-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,0,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,3,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,4,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,6,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,policy,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,0,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,3,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,4,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,5,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,6,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_cleanup_pencils,case4,unitree_z1_dual_arm_cleanup_pencils-vid200-fs4,unitree_z1_dual_arm_cleanup_pencils,world_model,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,policy,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,policy,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,policy,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,policy,4,0.95,1.0,1.0,45.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,policy,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,policy,6,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,world_model,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,world_model,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,world_model,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,world_model,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,world_model,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case1,unitree_z1_dual_arm_stackbox-vid5-fs4,unitree_z1_dual_arm_stackbox,world_model,6,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,policy,0,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,policy,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,policy,3,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,policy,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,policy,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,policy,6,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,world_model,0,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,world_model,2,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,world_model,3,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,world_model,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,world_model,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case2,unitree_z1_dual_arm_stackbox-vid15-fs4,unitree_z1_dual_arm_stackbox,world_model,6,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,policy,0,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,policy,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,policy,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,policy,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,policy,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,policy,6,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,world_model,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,world_model,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,world_model,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,world_model,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,world_model,5,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case3,unitree_z1_dual_arm_stackbox-vid25-fs4,unitree_z1_dual_arm_stackbox,world_model,6,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,policy,0,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,policy,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,policy,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,policy,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,policy,5,0.95,1.0,1.0,46.0,48.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,policy,6,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,world_model,0,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,world_model,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,world_model,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,world_model,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,world_model,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox,case4,unitree_z1_dual_arm_stackbox-vid35-fs4,unitree_z1_dual_arm_stackbox,world_model,6,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,0,0.95,1.0,1.0,46.0,45.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,2,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,4,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,5,0.95,1.0,1.0,45.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,6,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,7,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,8,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,9,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,policy,10,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,0,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,2,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,4,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,5,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,6,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,7,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,8,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,9,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case1,unitree_z1_dual_arm_stackbox_v2-vid5-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,10,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,3,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,4,0.95,1.0,1.0,45.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,5,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,7,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,9,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,policy,10,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,2,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,3,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,4,0.95,1.0,1.0,45.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,5,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,7,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,9,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case2,unitree_z1_dual_arm_stackbox_v2-vid15-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,10,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,2,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,4,0.95,1.0,1.0,45.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,5,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,7,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,9,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,policy,10,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,1,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,2,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,3,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,5,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,7,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,9,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case3,unitree_z1_dual_arm_stackbox_v2-vid25-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,10,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,1,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,2,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,4,0.95,1.0,1.0,45.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,7,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,9,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,policy,10,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,1,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,2,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,3,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,4,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,5,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,6,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,9,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_dual_arm_stackbox_v2,case4,unitree_z1_dual_arm_stackbox_v2-vid35-fs4,unitree_z1_dual_arm_stackbox_v2,world_model,10,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,1,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,2,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,3,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,4,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,5,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,6,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,7,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,8,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,9,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,10,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,policy,11,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,1,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,2,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,3,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,4,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,5,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,6,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,7,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,9,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,10,0.95,1.0,1.0,48.0,48.0,0.98,0.98
+unitree_z1_stackbox,case1,unitree_z1_stackbox-vid5-fs4,unitree_z1_stackbox,world_model,11,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,0,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,1,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,2,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,3,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,4,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,5,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,6,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,8,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,9,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,10,0.95,1.0,1.0,46.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,policy,11,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,0,0.95,1.0,1.0,47.0,46.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,1,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,2,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,3,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,4,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,5,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,6,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,9,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,10,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case2,unitree_z1_stackbox-vid15-fs4,unitree_z1_stackbox,world_model,11,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,1,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,2,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,3,0.95,1.0,1.0,46.0,48.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,5,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,6,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,9,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,10,0.95,1.0,1.0,45.0,45.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,policy,11,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,1,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,2,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,3,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,4,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,5,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,6,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,7,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,8,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,9,0.95,1.0,1.0,45.0,47.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,10,0.95,1.0,1.0,45.0,46.0,0.98,0.98
+unitree_z1_stackbox,case3,unitree_z1_stackbox-vid25-fs4,unitree_z1_stackbox,world_model,11,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,0,0.95,1.0,1.0,46.0,46.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,1,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,2,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,3,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,4,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,5,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,6,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,7,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,8,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,9,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,10,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,policy,11,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,0,0.95,1.0,1.0,47.0,46.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,1,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,2,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,3,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,4,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,5,0.95,1.0,1.0,47.0,47.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,6,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,7,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,8,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,9,0.95,1.0,1.0,46.0,47.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,10,0.95,1.0,1.0,47.0,48.0,0.98,0.98
+unitree_z1_stackbox,case4,unitree_z1_stackbox-vid35-fs4,unitree_z1_stackbox,world_model,11,0.95,1.0,1.0,47.0,48.0,0.98,0.98

sparse_head_simulation/report.md

@@ -0,0 +1,108 @@
+# Sparse Head Execution Simulation
+
+This report uses zero-order hold over logged stepwise action/state outputs.
+For a sparse scheme, the final output at step 49 is approximated by the most recent checkpoint output.
+
+## Experiment 1.1
+
+### dense
+
+- Head exec count: 50
+- Head compute saving ratio: 0.0000
+- Final action cosine vs dense: mean=1.0000, median=1.0000
+- Final state cosine vs dense: mean=1.0000, median=1.0000
+
+### sparse_10
+
+- Head exec count: 10
+- Head compute saving ratio: 0.8000
+- Final action cosine vs dense: mean=0.9341, median=0.9357
+- Final state cosine vs dense: mean=0.9130, median=0.9207
+
+### sparse_4
+
+- Head exec count: 4
+- Head compute saving ratio: 0.9200
+- Final action cosine vs dense: mean=1.0000, median=1.0000
+- Final state cosine vs dense: mean=1.0000, median=1.0000
+
+### sparse_5
+
+- Head exec count: 5
+- Head compute saving ratio: 0.9000
+- Final action cosine vs dense: mean=0.7747, median=0.7703
+- Final state cosine vs dense: mean=0.7463, median=0.7481
+
+### sparse_8
+
+- Head exec count: 8
+- Head compute saving ratio: 0.8400
+- Final action cosine vs dense: mean=1.0000, median=1.0000
+- Final state cosine vs dense: mean=1.0000, median=1.0000
+
+### tail_heavy_10
+
+- Head exec count: 10
+- Head compute saving ratio: 0.8000
+- Final action cosine vs dense: mean=1.0000, median=1.0000
+- Final state cosine vs dense: mean=1.0000, median=1.0000
+
+### tail_heavy_6
+
+- Head exec count: 6
+- Head compute saving ratio: 0.8800
+- Final action cosine vs dense: mean=1.0000, median=1.0000
+- Final state cosine vs dense: mean=1.0000, median=1.0000
+
+### tail_heavy_8
+
+- Head exec count: 8
+- Head compute saving ratio: 0.8400
+- Final action cosine vs dense: mean=1.0000, median=1.0000
+- Final state cosine vs dense: mean=1.0000, median=1.0000
+
+### tail_only_4
+
+- Head exec count: 4
+- Head compute saving ratio: 0.9200
+- Final action cosine vs dense: mean=1.0000, median=1.0000
+- Final state cosine vs dense: mean=1.0000, median=1.0000
+
+### tail_only_6
+
+- Head exec count: 6
+- Head compute saving ratio: 0.8800
+- Final action cosine vs dense: mean=1.0000, median=1.0000
+- Final state cosine vs dense: mean=1.0000, median=1.0000
+
+## Experiment 1.2
+
+- Target threshold: 0.95
+- Action earliest checkpoint mean: 46.0179
+- State earliest checkpoint mean: 46.5357
+- Unique min head steps needed for action: [1.0]
+- Unique min head steps needed for state: [1.0]
+
+## Experiment 1.3
+
+### budget_10_uniform_vs_tail
+
+- Tail better action share: 1.0000
+- Tail better state share: 1.0000
+- Mean tail-minus-uniform action cosine: 0.0659
+- Mean tail-minus-uniform state cosine: 0.0870
+
+### budget_4_uniform_vs_tail
+
+- Tail better action share: 0.0000
+- Tail better state share: 0.0000
+- Mean tail-minus-uniform action cosine: 0.0000
+- Mean tail-minus-uniform state cosine: 0.0000
+
+### budget_8_uniform_vs_tail
+
+- Tail better action share: 0.0000
+- Tail better state share: 0.0000
+- Mean tail-minus-uniform action cosine: 0.0000
+- Mean tail-minus-uniform state cosine: 0.0000
+

sparse_head_simulation/scheme_simulation_overall.csv

@@ -0,0 +1,22 @@
+scheme,pass_type,final_action_cosine_vs_dense,final_action_cosine_vs_dense,final_action_cosine_vs_dense,final_state_cosine_vs_dense,final_state_cosine_vs_dense,final_state_cosine_vs_dense,head_exec_count,head_exec_count,head_exec_count,head_compute_saving_ratio,head_compute_saving_ratio,head_compute_saving_ratio
+,,mean,median,std,mean,median,std,mean,median,std,mean,median,std
+dense,policy,1.000000006386212,1.0,8.6717418229413e-08,0.9999999939178934,1.0,9.826164033417286e-08,50.0,50.0,0.0,0.0,0.0,0.0
+dense,world_model,1.0000000069944226,1.0,9.373933947591167e-08,0.9999999936137881,1.0,9.102258232536112e-08,50.0,50.0,0.0,0.0,0.0,0.0
+sparse_10,policy,0.9357762607384701,0.9370943009853362,0.018183880133080282,0.9124557038350981,0.9198094010353088,0.03185505383051429,10.0,10.0,0.0,0.8,0.8,0.0
+sparse_10,world_model,0.9325069240769561,0.9341191649436951,0.0185309603092533,0.9134697318077087,0.9217162132263184,0.032954335242965455,10.0,10.0,0.0,0.8,0.8,0.0
+sparse_4,policy,1.000000006386212,1.0,8.6717418229413e-08,0.9999999939178934,1.0,9.826164033417286e-08,4.0,4.0,0.0,0.9200000000000002,0.92,0.0
+sparse_4,world_model,1.0000000069944226,1.0,9.373933947591167e-08,0.9999999936137881,1.0,9.102258232536112e-08,4.0,4.0,0.0,0.9200000000000002,0.92,0.0
+sparse_5,policy,0.7812822941614657,0.7787671983242035,0.03869909771001223,0.7465329766273499,0.7454181015491486,0.05109581488207182,5.0,5.0,0.0,0.9,0.9,0.0
+sparse_5,world_model,0.7681533368874569,0.7654141485691071,0.03844469190746414,0.7460347663383095,0.7492875158786774,0.05257641591339498,5.0,5.0,0.0,0.9,0.9,0.0
+sparse_8,policy,1.000000006386212,1.0,8.6717418229413e-08,0.9999999939178934,1.0,9.826164033417286e-08,8.0,8.0,0.0,0.84,0.84,0.0
+sparse_8,world_model,1.0000000069944226,1.0,9.373933947591167e-08,0.9999999936137881,1.0,9.102258232536112e-08,8.0,8.0,0.0,0.84,0.84,0.0
+tail_heavy_10,policy,1.000000006386212,1.0,8.6717418229413e-08,0.9999999939178934,1.0,9.826164033417286e-08,10.0,10.0,0.0,0.8,0.8,0.0
+tail_heavy_10,world_model,1.0000000069944226,1.0,9.373933947591167e-08,0.9999999936137881,1.0,9.102258232536112e-08,10.0,10.0,0.0,0.8,0.8,0.0
+tail_heavy_6,policy,1.000000006386212,1.0,8.6717418229413e-08,0.9999999939178934,1.0,9.826164033417286e-08,6.0,6.0,0.0,0.8799999999999999,0.88,0.0
+tail_heavy_6,world_model,1.0000000069944226,1.0,9.373933947591167e-08,0.9999999936137881,1.0,9.102258232536112e-08,6.0,6.0,0.0,0.8799999999999999,0.88,0.0
+tail_heavy_8,policy,1.000000006386212,1.0,8.6717418229413e-08,0.9999999939178934,1.0,9.826164033417286e-08,8.0,8.0,0.0,0.84,0.84,0.0
+tail_heavy_8,world_model,1.0000000069944226,1.0,9.373933947591167e-08,0.9999999936137881,1.0,9.102258232536112e-08,8.0,8.0,0.0,0.84,0.84,0.0
+tail_only_4,policy,1.000000006386212,1.0,8.6717418229413e-08,0.9999999939178934,1.0,9.826164033417286e-08,4.0,4.0,0.0,0.9200000000000002,0.92,0.0
+tail_only_4,world_model,1.0000000069944226,1.0,9.373933947591167e-08,0.9999999936137881,1.0,9.102258232536112e-08,4.0,4.0,0.0,0.9200000000000002,0.92,0.0
+tail_only_6,policy,1.000000006386212,1.0,8.6717418229413e-08,0.9999999939178934,1.0,9.826164033417286e-08,6.0,6.0,0.0,0.8799999999999999,0.88,0.0
+tail_only_6,world_model,1.0000000069944226,1.0,9.373933947591167e-08,0.9999999936137881,1.0,9.102258232536112e-08,6.0,6.0,0.0,0.8799999999999999,0.88,0.0

sparse_head_simulation/scheme_simulation_per_case.csv

@@ -0,0 +1,401 @@
+dataset,case,scheme,pass_type,final_action_cosine_vs_dense,final_state_cosine_vs_dense,head_exec_count,head_compute_saving_ratio
+unitree_g1_pack_camera,case1,dense,policy,1.000000005418604,1.0000000216744163,50.0,0.0
+unitree_g1_pack_camera,case1,dense,world_model,1.0000000162558123,0.9999999891627919,50.0,0.0
+unitree_g1_pack_camera,case1,sparse_10,policy,0.9458914236588911,0.9406864263794639,10.0,0.8
+unitree_g1_pack_camera,case1,sparse_10,world_model,0.943276660008864,0.94043137810447,10.0,0.8
+unitree_g1_pack_camera,case1,sparse_4,policy,1.000000005418604,1.0000000216744163,4.0,0.92
+unitree_g1_pack_camera,case1,sparse_4,world_model,1.0000000162558123,0.9999999891627919,4.0,0.92
+unitree_g1_pack_camera,case1,sparse_5,policy,0.7835968299345537,0.7882764935493469,5.0,0.9
+unitree_g1_pack_camera,case1,sparse_5,world_model,0.768740404735912,0.7820447520776228,5.0,0.9
+unitree_g1_pack_camera,case1,sparse_8,policy,1.000000005418604,1.0000000216744163,8.0,0.84
+unitree_g1_pack_camera,case1,sparse_8,world_model,1.0000000162558123,0.9999999891627919,8.0,0.84
+unitree_g1_pack_camera,case1,tail_heavy_10,policy,1.000000005418604,1.0000000216744163,10.0,0.8
+unitree_g1_pack_camera,case1,tail_heavy_10,world_model,1.0000000162558123,0.9999999891627919,10.0,0.8
+unitree_g1_pack_camera,case1,tail_heavy_6,policy,1.000000005418604,1.0000000216744163,6.0,0.88
+unitree_g1_pack_camera,case1,tail_heavy_6,world_model,1.0000000162558123,0.9999999891627919,6.0,0.88
+unitree_g1_pack_camera,case1,tail_heavy_8,policy,1.000000005418604,1.0000000216744163,8.0,0.84
+unitree_g1_pack_camera,case1,tail_heavy_8,world_model,1.0000000162558123,0.9999999891627919,8.0,0.84
+unitree_g1_pack_camera,case1,tail_only_4,policy,1.000000005418604,1.0000000216744163,4.0,0.92
+unitree_g1_pack_camera,case1,tail_only_4,world_model,1.0000000162558123,0.9999999891627919,4.0,0.92
+unitree_g1_pack_camera,case1,tail_only_6,policy,1.000000005418604,1.0000000216744163,6.0,0.88
+unitree_g1_pack_camera,case1,tail_only_6,world_model,1.0000000162558123,0.9999999891627919,6.0,0.88
+unitree_g1_pack_camera,case2,dense,policy,1.0,1.0000000108372082,50.0,0.0
+unitree_g1_pack_camera,case2,dense,world_model,1.0000000433488325,1.0,50.0,0.0
+unitree_g1_pack_camera,case2,sparse_10,policy,0.9434121738780629,0.9354661919853904,10.0,0.8
+unitree_g1_pack_camera,case2,sparse_10,world_model,0.9378488930788907,0.9352135224775835,10.0,0.8
+unitree_g1_pack_camera,case2,sparse_4,policy,1.0,1.0000000108372082,4.0,0.92
+unitree_g1_pack_camera,case2,sparse_4,world_model,1.0000000433488325,1.0,4.0,0.92
+unitree_g1_pack_camera,case2,sparse_5,policy,0.7797025496309454,0.7743967554785989,5.0,0.9
+unitree_g1_pack_camera,case2,sparse_5,world_model,0.7510665926066312,0.7674615762450478,5.0,0.9
+unitree_g1_pack_camera,case2,sparse_8,policy,1.0,1.0000000108372082,8.0,0.84
+unitree_g1_pack_camera,case2,sparse_8,world_model,1.0000000433488325,1.0,8.0,0.84
+unitree_g1_pack_camera,case2,tail_heavy_10,policy,1.0,1.0000000108372082,10.0,0.8
+unitree_g1_pack_camera,case2,tail_heavy_10,world_model,1.0000000433488325,1.0,10.0,0.8
+unitree_g1_pack_camera,case2,tail_heavy_6,policy,1.0,1.0000000108372082,6.0,0.88
+unitree_g1_pack_camera,case2,tail_heavy_6,world_model,1.0000000433488325,1.0,6.0,0.88
+unitree_g1_pack_camera,case2,tail_heavy_8,policy,1.0,1.0000000108372082,8.0,0.84
+unitree_g1_pack_camera,case2,tail_heavy_8,world_model,1.0000000433488325,1.0,8.0,0.84
+unitree_g1_pack_camera,case2,tail_only_4,policy,1.0,1.0000000108372082,4.0,0.92
+unitree_g1_pack_camera,case2,tail_only_4,world_model,1.0000000433488325,1.0,4.0,0.92
+unitree_g1_pack_camera,case2,tail_only_6,policy,1.0,1.0000000108372082,6.0,0.88
+unitree_g1_pack_camera,case2,tail_only_6,world_model,1.0000000433488325,1.0,6.0,0.88
+unitree_g1_pack_camera,case3,dense,policy,0.9999999729069796,0.9999999945813959,50.0,0.0
+unitree_g1_pack_camera,case3,dense,world_model,0.9999999891627919,0.9999999729069796,50.0,0.0
+unitree_g1_pack_camera,case3,sparse_10,policy,0.9479800787839022,0.9408404772931879,10.0,0.8
+unitree_g1_pack_camera,case3,sparse_10,world_model,0.9443203156644647,0.9426048723134127,10.0,0.8
+unitree_g1_pack_camera,case3,sparse_4,policy,0.9999999729069796,0.9999999945813959,4.0,0.92
+unitree_g1_pack_camera,case3,sparse_4,world_model,0.9999999891627919,0.9999999729069796,4.0,0.92
+unitree_g1_pack_camera,case3,sparse_5,policy,0.7867171547629617,0.7888143712824042,5.0,0.9
+unitree_g1_pack_camera,case3,sparse_5,world_model,0.7702723687345331,0.7856205864386125,5.0,0.9
+unitree_g1_pack_camera,case3,sparse_8,policy,0.9999999729069796,0.9999999945813959,8.0,0.84
+unitree_g1_pack_camera,case3,sparse_8,world_model,0.9999999891627919,0.9999999729069796,8.0,0.84
+unitree_g1_pack_camera,case3,tail_heavy_10,policy,0.9999999729069796,0.9999999945813959,10.0,0.8
+unitree_g1_pack_camera,case3,tail_heavy_10,world_model,0.9999999891627919,0.9999999729069796,10.0,0.8
+unitree_g1_pack_camera,case3,tail_heavy_6,policy,0.9999999729069796,0.9999999945813959,6.0,0.88
+unitree_g1_pack_camera,case3,tail_heavy_6,world_model,0.9999999891627919,0.9999999729069796,6.0,0.88
+unitree_g1_pack_camera,case3,tail_heavy_8,policy,0.9999999729069796,0.9999999945813959,8.0,0.84
+unitree_g1_pack_camera,case3,tail_heavy_8,world_model,0.9999999891627919,0.9999999729069796,8.0,0.84
+unitree_g1_pack_camera,case3,tail_only_4,policy,0.9999999729069796,0.9999999945813959,4.0,0.92
+unitree_g1_pack_camera,case3,tail_only_4,world_model,0.9999999891627919,0.9999999729069796,4.0,0.92
+unitree_g1_pack_camera,case3,tail_only_6,policy,0.9999999729069796,0.9999999945813959,6.0,0.88
+unitree_g1_pack_camera,case3,tail_only_6,world_model,0.9999999891627919,0.9999999729069796,6.0,0.88
+unitree_g1_pack_camera,case4,dense,policy,0.9999999783255837,1.000000005418604,50.0,0.0
+unitree_g1_pack_camera,case4,dense,world_model,1.000000005418604,1.0000000162558123,50.0,0.0
+unitree_g1_pack_camera,case4,sparse_10,policy,0.9499884627082131,0.9411461407488043,10.0,0.8
+unitree_g1_pack_camera,case4,sparse_10,world_model,0.9460353851318359,0.9442069042812694,10.0,0.8
+unitree_g1_pack_camera,case4,sparse_4,policy,0.9999999783255837,1.000000005418604,4.0,0.92
+unitree_g1_pack_camera,case4,sparse_4,world_model,1.000000005418604,1.0000000162558123,4.0,0.92
+unitree_g1_pack_camera,case4,sparse_5,policy,0.797451143915003,0.786096372387626,5.0,0.9
+unitree_g1_pack_camera,case4,sparse_5,world_model,0.776059172370217,0.7922905737703497,5.0,0.9
+unitree_g1_pack_camera,case4,sparse_8,policy,0.9999999783255837,1.000000005418604,8.0,0.84
+unitree_g1_pack_camera,case4,sparse_8,world_model,1.000000005418604,1.0000000162558123,8.0,0.84
+unitree_g1_pack_camera,case4,tail_heavy_10,policy,0.9999999783255837,1.000000005418604,10.0,0.8
+unitree_g1_pack_camera,case4,tail_heavy_10,world_model,1.000000005418604,1.0000000162558123,10.0,0.8
+unitree_g1_pack_camera,case4,tail_heavy_6,policy,0.9999999783255837,1.000000005418604,6.0,0.88
+unitree_g1_pack_camera,case4,tail_heavy_6,world_model,1.000000005418604,1.0000000162558123,6.0,0.88
+unitree_g1_pack_camera,case4,tail_heavy_8,policy,0.9999999783255837,1.000000005418604,8.0,0.84
+unitree_g1_pack_camera,case4,tail_heavy_8,world_model,1.000000005418604,1.0000000162558123,8.0,0.84
+unitree_g1_pack_camera,case4,tail_only_4,policy,0.9999999783255837,1.000000005418604,4.0,0.92
+unitree_g1_pack_camera,case4,tail_only_4,world_model,1.000000005418604,1.0000000162558123,4.0,0.92
+unitree_g1_pack_camera,case4,tail_only_6,policy,0.9999999783255837,1.000000005418604,6.0,0.88
+unitree_g1_pack_camera,case4,tail_only_6,world_model,1.000000005418604,1.0000000162558123,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case1,dense,policy,1.0000000074505806,0.9999999850988388,50.0,0.0
+unitree_z1_dual_arm_cleanup_pencils,case1,dense,world_model,0.9999999925494194,1.000000037252903,50.0,0.0
+unitree_z1_dual_arm_cleanup_pencils,case1,sparse_10,policy,0.9469437971711159,0.9320246428251266,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case1,sparse_10,world_model,0.9465975016355515,0.9307577759027481,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case1,sparse_4,policy,1.0000000074505806,0.9999999850988388,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case1,sparse_4,world_model,0.9999999925494194,1.000000037252903,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case1,sparse_5,policy,0.8056009188294411,0.7804089784622192,5.0,0.9
+unitree_z1_dual_arm_cleanup_pencils,case1,sparse_5,world_model,0.800347164273262,0.7744432389736176,5.0,0.9
+unitree_z1_dual_arm_cleanup_pencils,case1,sparse_8,policy,1.0000000074505806,0.9999999850988388,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case1,sparse_8,world_model,0.9999999925494194,1.000000037252903,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_heavy_10,policy,1.0000000074505806,0.9999999850988388,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_heavy_10,world_model,0.9999999925494194,1.000000037252903,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_heavy_6,policy,1.0000000074505806,0.9999999850988388,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_heavy_6,world_model,0.9999999925494194,1.000000037252903,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_heavy_8,policy,1.0000000074505806,0.9999999850988388,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_heavy_8,world_model,0.9999999925494194,1.000000037252903,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_only_4,policy,1.0000000074505806,0.9999999850988388,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_only_4,world_model,0.9999999925494194,1.000000037252903,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_only_6,policy,1.0000000074505806,0.9999999850988388,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case1,tail_only_6,world_model,0.9999999925494194,1.000000037252903,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case2,dense,policy,1.0000000223517418,0.999999962747097,50.0,0.0
+unitree_z1_dual_arm_cleanup_pencils,case2,dense,world_model,1.0000000298023224,0.9999999850988388,50.0,0.0
+unitree_z1_dual_arm_cleanup_pencils,case2,sparse_10,policy,0.9463301375508308,0.9295393601059914,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case2,sparse_10,world_model,0.9458944350481033,0.9283439442515373,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case2,sparse_4,policy,1.0000000223517418,0.999999962747097,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case2,sparse_4,world_model,1.0000000298023224,0.9999999850988388,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case2,sparse_5,policy,0.809041365981102,0.7754942625761032,5.0,0.9
+unitree_z1_dual_arm_cleanup_pencils,case2,sparse_5,world_model,0.805028922855854,0.7709039524197578,5.0,0.9
+unitree_z1_dual_arm_cleanup_pencils,case2,sparse_8,policy,1.0000000223517418,0.999999962747097,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case2,sparse_8,world_model,1.0000000298023224,0.9999999850988388,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_heavy_10,policy,1.0000000223517418,0.999999962747097,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_heavy_10,world_model,1.0000000298023224,0.9999999850988388,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_heavy_6,policy,1.0000000223517418,0.999999962747097,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_heavy_6,world_model,1.0000000298023224,0.9999999850988388,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_heavy_8,policy,1.0000000223517418,0.999999962747097,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_heavy_8,world_model,1.0000000298023224,0.9999999850988388,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_only_4,policy,1.0000000223517418,0.999999962747097,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_only_4,world_model,1.0000000298023224,0.9999999850988388,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_only_6,policy,1.0000000223517418,0.999999962747097,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case2,tail_only_6,world_model,1.0000000298023224,0.9999999850988388,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case3,dense,policy,0.9999999925494194,0.9999999850988388,50.0,0.0
+unitree_z1_dual_arm_cleanup_pencils,case3,dense,world_model,0.9999999701976776,0.999999925494194,50.0,0.0
+unitree_z1_dual_arm_cleanup_pencils,case3,sparse_10,policy,0.9426491186022758,0.9190819412469864,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case3,sparse_10,world_model,0.939857967197895,0.9199994206428528,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case3,sparse_4,policy,0.9999999925494194,0.9999999850988388,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case3,sparse_4,world_model,0.9999999701976776,0.999999925494194,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case3,sparse_5,policy,0.7971489876508713,0.7457228153944016,5.0,0.9
+unitree_z1_dual_arm_cleanup_pencils,case3,sparse_5,world_model,0.7816280424594879,0.7513428404927254,5.0,0.9
+unitree_z1_dual_arm_cleanup_pencils,case3,sparse_8,policy,0.9999999925494194,0.9999999850988388,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case3,sparse_8,world_model,0.9999999701976776,0.999999925494194,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_heavy_10,policy,0.9999999925494194,0.9999999850988388,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_heavy_10,world_model,0.9999999701976776,0.999999925494194,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_heavy_6,policy,0.9999999925494194,0.9999999850988388,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_heavy_6,world_model,0.9999999701976776,0.999999925494194,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_heavy_8,policy,0.9999999925494194,0.9999999850988388,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_heavy_8,world_model,0.9999999701976776,0.999999925494194,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_only_4,policy,0.9999999925494194,0.9999999850988388,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_only_4,world_model,0.9999999701976776,0.999999925494194,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_only_6,policy,0.9999999925494194,0.9999999850988388,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case3,tail_only_6,world_model,0.9999999701976776,0.999999925494194,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case4,dense,policy,1.0,1.0,50.0,0.0
+unitree_z1_dual_arm_cleanup_pencils,case4,dense,world_model,1.0,0.9999999105930328,50.0,0.0
+unitree_z1_dual_arm_cleanup_pencils,case4,sparse_10,policy,0.9463553726673126,0.9252839460968971,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case4,sparse_10,world_model,0.9459596052765846,0.9299178943037987,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case4,sparse_4,policy,1.0,1.0,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case4,sparse_4,world_model,1.0,0.9999999105930328,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case4,sparse_5,policy,0.8052143007516861,0.7575837001204491,5.0,0.9
+unitree_z1_dual_arm_cleanup_pencils,case4,sparse_5,world_model,0.7977565228939056,0.7714092582464218,5.0,0.9
+unitree_z1_dual_arm_cleanup_pencils,case4,sparse_8,policy,1.0,1.0,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case4,sparse_8,world_model,1.0,0.9999999105930328,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_heavy_10,policy,1.0,1.0,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_heavy_10,world_model,1.0,0.9999999105930328,10.0,0.8
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_heavy_6,policy,1.0,1.0,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_heavy_6,world_model,1.0,0.9999999105930328,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_heavy_8,policy,1.0,1.0,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_heavy_8,world_model,1.0,0.9999999105930328,8.0,0.84
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_only_4,policy,1.0,1.0,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_only_4,world_model,1.0,0.9999999105930328,4.0,0.92
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_only_6,policy,1.0,1.0,6.0,0.88
+unitree_z1_dual_arm_cleanup_pencils,case4,tail_only_6,world_model,1.0,0.9999999105930328,6.0,0.88
+unitree_z1_dual_arm_stackbox,case1,dense,policy,1.0000000170298986,1.0,50.0,0.0
+unitree_z1_dual_arm_stackbox,case1,dense,world_model,1.0000000255448478,1.0,50.0,0.0
+unitree_z1_dual_arm_stackbox,case1,sparse_10,policy,0.9336806450571332,0.9031614576067243,10.0,0.8
+unitree_z1_dual_arm_stackbox,case1,sparse_10,world_model,0.9270949619156974,0.9099422693252563,10.0,0.8
+unitree_z1_dual_arm_stackbox,case1,sparse_4,policy,1.0000000170298986,1.0,4.0,0.92
+unitree_z1_dual_arm_stackbox,case1,sparse_4,world_model,1.0000000255448478,1.0,4.0,0.92
+unitree_z1_dual_arm_stackbox,case1,sparse_5,policy,0.7690511431012835,0.7218933616365705,5.0,0.9
+unitree_z1_dual_arm_stackbox,case1,sparse_5,world_model,0.7506173253059387,0.7309779609952655,5.0,0.9
+unitree_z1_dual_arm_stackbox,case1,sparse_8,policy,1.0000000170298986,1.0,8.0,0.84
+unitree_z1_dual_arm_stackbox,case1,sparse_8,world_model,1.0000000255448478,1.0,8.0,0.84
+unitree_z1_dual_arm_stackbox,case1,tail_heavy_10,policy,1.0000000170298986,1.0,10.0,0.8
+unitree_z1_dual_arm_stackbox,case1,tail_heavy_10,world_model,1.0000000255448478,1.0,10.0,0.8
+unitree_z1_dual_arm_stackbox,case1,tail_heavy_6,policy,1.0000000170298986,1.0,6.0,0.88
+unitree_z1_dual_arm_stackbox,case1,tail_heavy_6,world_model,1.0000000255448478,1.0,6.0,0.88
+unitree_z1_dual_arm_stackbox,case1,tail_heavy_8,policy,1.0000000170298986,1.0,8.0,0.84
+unitree_z1_dual_arm_stackbox,case1,tail_heavy_8,world_model,1.0000000255448478,1.0,8.0,0.84
+unitree_z1_dual_arm_stackbox,case1,tail_only_4,policy,1.0000000170298986,1.0,4.0,0.92
+unitree_z1_dual_arm_stackbox,case1,tail_only_4,world_model,1.0000000255448478,1.0,4.0,0.92
+unitree_z1_dual_arm_stackbox,case1,tail_only_6,policy,1.0000000170298986,1.0,6.0,0.88
+unitree_z1_dual_arm_stackbox,case1,tail_only_6,world_model,1.0000000255448478,1.0,6.0,0.88
+unitree_z1_dual_arm_stackbox,case2,dense,policy,0.9999999914850507,0.9999998978206089,50.0,0.0
+unitree_z1_dual_arm_stackbox,case2,dense,world_model,1.000000068119594,1.0000000510896956,50.0,0.0
+unitree_z1_dual_arm_stackbox,case2,sparse_10,policy,0.9294669883591788,0.899870480809893,10.0,0.8
+unitree_z1_dual_arm_stackbox,case2,sparse_10,world_model,0.9253293957029071,0.9051141398293632,10.0,0.8
+unitree_z1_dual_arm_stackbox,case2,sparse_4,policy,0.9999999914850507,0.9999998978206089,4.0,0.92
+unitree_z1_dual_arm_stackbox,case2,sparse_4,world_model,1.000000068119594,1.0000000510896956,4.0,0.92
+unitree_z1_dual_arm_stackbox,case2,sparse_5,policy,0.7612549066543579,0.7151832665715899,5.0,0.9
+unitree_z1_dual_arm_stackbox,case2,sparse_5,world_model,0.744773634842464,0.7192469154085431,5.0,0.9
+unitree_z1_dual_arm_stackbox,case2,sparse_8,policy,0.9999999914850507,0.9999998978206089,8.0,0.84
+unitree_z1_dual_arm_stackbox,case2,sparse_8,world_model,1.000000068119594,1.0000000510896956,8.0,0.84
+unitree_z1_dual_arm_stackbox,case2,tail_heavy_10,policy,0.9999999914850507,0.9999998978206089,10.0,0.8
+unitree_z1_dual_arm_stackbox,case2,tail_heavy_10,world_model,1.000000068119594,1.0000000510896956,10.0,0.8
+unitree_z1_dual_arm_stackbox,case2,tail_heavy_6,policy,0.9999999914850507,0.9999998978206089,6.0,0.88
+unitree_z1_dual_arm_stackbox,case2,tail_heavy_6,world_model,1.000000068119594,1.0000000510896956,6.0,0.88
+unitree_z1_dual_arm_stackbox,case2,tail_heavy_8,policy,0.9999999914850507,0.9999998978206089,8.0,0.84
+unitree_z1_dual_arm_stackbox,case2,tail_heavy_8,world_model,1.000000068119594,1.0000000510896956,8.0,0.84
+unitree_z1_dual_arm_stackbox,case2,tail_only_4,policy,0.9999999914850507,0.9999998978206089,4.0,0.92
+unitree_z1_dual_arm_stackbox,case2,tail_only_4,world_model,1.000000068119594,1.0000000510896956,4.0,0.92
+unitree_z1_dual_arm_stackbox,case2,tail_only_6,policy,0.9999999914850507,0.9999998978206089,6.0,0.88
+unitree_z1_dual_arm_stackbox,case2,tail_only_6,world_model,1.000000068119594,1.0000000510896956,6.0,0.88
+unitree_z1_dual_arm_stackbox,case3,dense,policy,1.0000000255448478,1.0000000170298986,50.0,0.0
+unitree_z1_dual_arm_stackbox,case3,dense,world_model,0.9999999914850507,0.9999999744551522,50.0,0.0
+unitree_z1_dual_arm_stackbox,case3,sparse_10,policy,0.9321104969297137,0.8938331604003906,10.0,0.8
+unitree_z1_dual_arm_stackbox,case3,sparse_10,world_model,0.928542605468205,0.9024338807378497,10.0,0.8
+unitree_z1_dual_arm_stackbox,case3,sparse_4,policy,1.0000000255448478,1.0000000170298986,4.0,0.92
+unitree_z1_dual_arm_stackbox,case3,sparse_4,world_model,0.9999999914850507,0.9999999744551522,4.0,0.92
+unitree_z1_dual_arm_stackbox,case3,sparse_5,policy,0.7596460665975299,0.702316667352404,5.0,0.9
+unitree_z1_dual_arm_stackbox,case3,sparse_5,world_model,0.7449179036276681,0.7157356142997742,5.0,0.9
+unitree_z1_dual_arm_stackbox,case3,sparse_8,policy,1.0000000255448478,1.0000000170298986,8.0,0.84
+unitree_z1_dual_arm_stackbox,case3,sparse_8,world_model,0.9999999914850507,0.9999999744551522,8.0,0.84
+unitree_z1_dual_arm_stackbox,case3,tail_heavy_10,policy,1.0000000255448478,1.0000000170298986,10.0,0.8
+unitree_z1_dual_arm_stackbox,case3,tail_heavy_10,world_model,0.9999999914850507,0.9999999744551522,10.0,0.8
+unitree_z1_dual_arm_stackbox,case3,tail_heavy_6,policy,1.0000000255448478,1.0000000170298986,6.0,0.88
+unitree_z1_dual_arm_stackbox,case3,tail_heavy_6,world_model,0.9999999914850507,0.9999999744551522,6.0,0.88
+unitree_z1_dual_arm_stackbox,case3,tail_heavy_8,policy,1.0000000255448478,1.0000000170298986,8.0,0.84
+unitree_z1_dual_arm_stackbox,case3,tail_heavy_8,world_model,0.9999999914850507,0.9999999744551522,8.0,0.84
+unitree_z1_dual_arm_stackbox,case3,tail_only_4,policy,1.0000000255448478,1.0000000170298986,4.0,0.92
+unitree_z1_dual_arm_stackbox,case3,tail_only_4,world_model,0.9999999914850507,0.9999999744551522,4.0,0.92
+unitree_z1_dual_arm_stackbox,case3,tail_only_6,policy,1.0000000255448478,1.0000000170298986,6.0,0.88
+unitree_z1_dual_arm_stackbox,case3,tail_only_6,world_model,0.9999999914850507,0.9999999744551522,6.0,0.88
+unitree_z1_dual_arm_stackbox,case4,dense,policy,1.0,0.9999999489103045,50.0,0.0
+unitree_z1_dual_arm_stackbox,case4,dense,world_model,1.0,0.9999999233654567,50.0,0.0
+unitree_z1_dual_arm_stackbox,case4,sparse_10,policy,0.9348850761141095,0.895094633102417,10.0,0.8
+unitree_z1_dual_arm_stackbox,case4,sparse_10,world_model,0.9304544499942234,0.9077311754226685,10.0,0.8
+unitree_z1_dual_arm_stackbox,case4,sparse_4,policy,1.0,0.9999999489103045,4.0,0.92
+unitree_z1_dual_arm_stackbox,case4,sparse_4,world_model,1.0,0.9999999233654567,4.0,0.92
+unitree_z1_dual_arm_stackbox,case4,sparse_5,policy,0.7730573671204704,0.7030913659504482,5.0,0.9
+unitree_z1_dual_arm_stackbox,case4,sparse_5,world_model,0.751270455973489,0.7207447120121547,5.0,0.9
+unitree_z1_dual_arm_stackbox,case4,sparse_8,policy,1.0,0.9999999489103045,8.0,0.84
+unitree_z1_dual_arm_stackbox,case4,sparse_8,world_model,1.0,0.9999999233654567,8.0,0.84
+unitree_z1_dual_arm_stackbox,case4,tail_heavy_10,policy,1.0,0.9999999489103045,10.0,0.8
+unitree_z1_dual_arm_stackbox,case4,tail_heavy_10,world_model,1.0,0.9999999233654567,10.0,0.8
+unitree_z1_dual_arm_stackbox,case4,tail_heavy_6,policy,1.0,0.9999999489103045,6.0,0.88
+unitree_z1_dual_arm_stackbox,case4,tail_heavy_6,world_model,1.0,0.9999999233654567,6.0,0.88
+unitree_z1_dual_arm_stackbox,case4,tail_heavy_8,policy,1.0,0.9999999489103045,8.0,0.84
+unitree_z1_dual_arm_stackbox,case4,tail_heavy_8,world_model,1.0,0.9999999233654567,8.0,0.84
+unitree_z1_dual_arm_stackbox,case4,tail_only_4,policy,1.0,0.9999999489103045,4.0,0.92
+unitree_z1_dual_arm_stackbox,case4,tail_only_4,world_model,1.0,0.9999999233654567,4.0,0.92
+unitree_z1_dual_arm_stackbox,case4,tail_only_6,policy,1.0,0.9999999489103045,6.0,0.88
+unitree_z1_dual_arm_stackbox,case4,tail_only_6,world_model,1.0,0.9999999233654567,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case1,dense,policy,1.0000000162558123,0.9999999674883756,50.0,0.0
+unitree_z1_dual_arm_stackbox_v2,case1,dense,world_model,0.9999999729069796,1.0,50.0,0.0
+unitree_z1_dual_arm_stackbox_v2,case1,sparse_10,policy,0.9435484842820601,0.928479487245733,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case1,sparse_10,world_model,0.9468015432357788,0.9367762858217413,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case1,sparse_4,policy,1.0000000162558123,0.9999999674883756,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case1,sparse_4,world_model,0.9999999729069796,1.0,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case1,sparse_5,policy,0.7936065901409496,0.7711024501106956,5.0,0.9
+unitree_z1_dual_arm_stackbox_v2,case1,sparse_5,world_model,0.8046409541910345,0.7889893271706321,5.0,0.9
+unitree_z1_dual_arm_stackbox_v2,case1,sparse_8,policy,1.0000000162558123,0.9999999674883756,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case1,sparse_8,world_model,0.9999999729069796,1.0,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case1,tail_heavy_10,policy,1.0000000162558123,0.9999999674883756,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case1,tail_heavy_10,world_model,0.9999999729069796,1.0,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case1,tail_heavy_6,policy,1.0000000162558123,0.9999999674883756,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case1,tail_heavy_6,world_model,0.9999999729069796,1.0,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case1,tail_heavy_8,policy,1.0000000162558123,0.9999999674883756,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case1,tail_heavy_8,world_model,0.9999999729069796,1.0,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case1,tail_only_4,policy,1.0000000162558123,0.9999999674883756,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case1,tail_only_4,world_model,0.9999999729069796,1.0,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case1,tail_only_6,policy,1.0000000162558123,0.9999999674883756,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case1,tail_only_6,world_model,0.9999999729069796,1.0,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case2,dense,policy,1.0,1.0000000433488325,50.0,0.0
+unitree_z1_dual_arm_stackbox_v2,case2,dense,world_model,1.0000000270930203,0.9999999945813959,50.0,0.0
+unitree_z1_dual_arm_stackbox_v2,case2,sparse_10,policy,0.939013806256381,0.9207526391202753,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case2,sparse_10,world_model,0.9369614882902666,0.9168761806054548,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case2,sparse_4,policy,1.0,1.0000000433488325,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case2,sparse_4,world_model,1.0000000270930203,0.9999999945813959,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case2,sparse_5,policy,0.7819729284806685,0.755226807160811,5.0,0.9
+unitree_z1_dual_arm_stackbox_v2,case2,sparse_5,world_model,0.7765740643848072,0.743860358541662,5.0,0.9
+unitree_z1_dual_arm_stackbox_v2,case2,sparse_8,policy,1.0,1.0000000433488325,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case2,sparse_8,world_model,1.0000000270930203,0.9999999945813959,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case2,tail_heavy_10,policy,1.0,1.0000000433488325,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case2,tail_heavy_10,world_model,1.0000000270930203,0.9999999945813959,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case2,tail_heavy_6,policy,1.0,1.0000000433488325,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case2,tail_heavy_6,world_model,1.0000000270930203,0.9999999945813959,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case2,tail_heavy_8,policy,1.0,1.0000000433488325,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case2,tail_heavy_8,world_model,1.0000000270930203,0.9999999945813959,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case2,tail_only_4,policy,1.0,1.0000000433488325,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case2,tail_only_4,world_model,1.0000000270930203,0.9999999945813959,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case2,tail_only_6,policy,1.0,1.0000000433488325,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case2,tail_only_6,world_model,1.0000000270930203,0.9999999945813959,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case3,dense,policy,1.0000000162558123,0.9999999620697715,50.0,0.0
+unitree_z1_dual_arm_stackbox_v2,case3,dense,world_model,0.9999999837441877,1.0000000108372082,50.0,0.0
+unitree_z1_dual_arm_stackbox_v2,case3,sparse_10,policy,0.9378652030771429,0.9162919467145746,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case3,sparse_10,world_model,0.9346397735855796,0.9093327955766157,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case3,sparse_4,policy,1.0000000162558123,0.9999999620697715,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case3,sparse_4,world_model,0.9999999837441877,1.0000000108372082,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case3,sparse_5,policy,0.7820485776120966,0.7449031526392157,5.0,0.9
+unitree_z1_dual_arm_stackbox_v2,case3,sparse_5,world_model,0.7718150182203813,0.7291300730271773,5.0,0.9
+unitree_z1_dual_arm_stackbox_v2,case3,sparse_8,policy,1.0000000162558123,0.9999999620697715,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case3,sparse_8,world_model,0.9999999837441877,1.0000000108372082,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case3,tail_heavy_10,policy,1.0000000162558123,0.9999999620697715,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case3,tail_heavy_10,world_model,0.9999999837441877,1.0000000108372082,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case3,tail_heavy_6,policy,1.0000000162558123,0.9999999620697715,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case3,tail_heavy_6,world_model,0.9999999837441877,1.0000000108372082,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case3,tail_heavy_8,policy,1.0000000162558123,0.9999999620697715,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case3,tail_heavy_8,world_model,0.9999999837441877,1.0000000108372082,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case3,tail_only_4,policy,1.0000000162558123,0.9999999620697715,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case3,tail_only_4,world_model,0.9999999837441877,1.0000000108372082,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case3,tail_only_6,policy,1.0000000162558123,0.9999999620697715,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case3,tail_only_6,world_model,0.9999999837441877,1.0000000108372082,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case4,dense,policy,1.0000000270930203,0.9999999512325634,50.0,0.0
+unitree_z1_dual_arm_stackbox_v2,case4,dense,world_model,1.000000005418604,0.9999999783255837,50.0,0.0
+unitree_z1_dual_arm_stackbox_v2,case4,sparse_10,policy,0.9357004165649414,0.9140155152841047,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case4,sparse_10,world_model,0.9342409697445956,0.9105181964960966,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case4,sparse_4,policy,1.0000000270930203,0.9999999512325634,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case4,sparse_4,world_model,1.000000005418604,0.9999999783255837,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case4,sparse_5,policy,0.7757082148031755,0.7411849932237105,5.0,0.9
+unitree_z1_dual_arm_stackbox_v2,case4,sparse_5,world_model,0.7706848870624196,0.7300773588093844,5.0,0.9
+unitree_z1_dual_arm_stackbox_v2,case4,sparse_8,policy,1.0000000270930203,0.9999999512325634,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case4,sparse_8,world_model,1.000000005418604,0.9999999783255837,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case4,tail_heavy_10,policy,1.0000000270930203,0.9999999512325634,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case4,tail_heavy_10,world_model,1.000000005418604,0.9999999783255837,10.0,0.8
+unitree_z1_dual_arm_stackbox_v2,case4,tail_heavy_6,policy,1.0000000270930203,0.9999999512325634,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case4,tail_heavy_6,world_model,1.000000005418604,0.9999999783255837,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case4,tail_heavy_8,policy,1.0000000270930203,0.9999999512325634,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case4,tail_heavy_8,world_model,1.000000005418604,0.9999999783255837,8.0,0.84
+unitree_z1_dual_arm_stackbox_v2,case4,tail_only_4,policy,1.0000000270930203,0.9999999512325634,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case4,tail_only_4,world_model,1.000000005418604,0.9999999783255837,4.0,0.92
+unitree_z1_dual_arm_stackbox_v2,case4,tail_only_6,policy,1.0000000270930203,0.9999999512325634,6.0,0.88
+unitree_z1_dual_arm_stackbox_v2,case4,tail_only_6,world_model,1.000000005418604,0.9999999783255837,6.0,0.88
+unitree_z1_stackbox,case1,dense,policy,0.9999999602635702,1.0000000049670537,50.0,0.0
+unitree_z1_stackbox,case1,dense,world_model,0.9999999701976776,1.0000000149011612,50.0,0.0
+unitree_z1_stackbox,case1,sparse_10,policy,0.9128866543372472,0.8721721917390823,10.0,0.8000000000000002
+unitree_z1_stackbox,case1,sparse_10,world_model,0.9075939754645029,0.8709724992513657,10.0,0.8000000000000002
+unitree_z1_stackbox,case1,sparse_4,policy,0.9999999602635702,1.0000000049670537,4.0,0.92
+unitree_z1_stackbox,case1,sparse_4,world_model,0.9999999701976776,1.0000000149011612,4.0,0.92
+unitree_z1_stackbox,case1,sparse_5,policy,0.7592518925666809,0.7061114360888799,5.0,0.9
+unitree_z1_stackbox,case1,sparse_5,world_model,0.7442370106776556,0.7053270240624746,5.0,0.9
+unitree_z1_stackbox,case1,sparse_8,policy,0.9999999602635702,1.0000000049670537,8.0,0.84
+unitree_z1_stackbox,case1,sparse_8,world_model,0.9999999701976776,1.0000000149011612,8.0,0.84
+unitree_z1_stackbox,case1,tail_heavy_10,policy,0.9999999602635702,1.0000000049670537,10.0,0.8000000000000002
+unitree_z1_stackbox,case1,tail_heavy_10,world_model,0.9999999701976776,1.0000000149011612,10.0,0.8000000000000002
+unitree_z1_stackbox,case1,tail_heavy_6,policy,0.9999999602635702,1.0000000049670537,6.0,0.88
+unitree_z1_stackbox,case1,tail_heavy_6,world_model,0.9999999701976776,1.0000000149011612,6.0,0.88
+unitree_z1_stackbox,case1,tail_heavy_8,policy,0.9999999602635702,1.0000000049670537,8.0,0.84
+unitree_z1_stackbox,case1,tail_heavy_8,world_model,0.9999999701976776,1.0000000149011612,8.0,0.84
+unitree_z1_stackbox,case1,tail_only_4,policy,0.9999999602635702,1.0000000049670537,4.0,0.92
+unitree_z1_stackbox,case1,tail_only_4,world_model,0.9999999701976776,1.0000000149011612,4.0,0.92
+unitree_z1_stackbox,case1,tail_only_6,policy,0.9999999602635702,1.0000000049670537,6.0,0.88
+unitree_z1_stackbox,case1,tail_only_6,world_model,0.9999999701976776,1.0000000149011612,6.0,0.88
+unitree_z1_stackbox,case2,dense,policy,1.0000000099341075,1.000000034769376,50.0,0.0
+unitree_z1_stackbox,case2,dense,world_model,1.0000000198682149,1.0000000149011612,50.0,0.0
+unitree_z1_stackbox,case2,sparse_10,policy,0.9204017321268717,0.8808373361825943,10.0,0.8000000000000002
+unitree_z1_stackbox,case2,sparse_10,world_model,0.9124102195103964,0.8839416752258936,10.0,0.8000000000000002
+unitree_z1_stackbox,case2,sparse_4,policy,1.0000000099341075,1.000000034769376,4.0,0.92
+unitree_z1_stackbox,case2,sparse_4,world_model,1.0000000198682149,1.0000000149011612,4.0,0.92
+unitree_z1_stackbox,case2,sparse_5,policy,0.7721244891484579,0.7142822792132696,5.0,0.9
+unitree_z1_stackbox,case2,sparse_5,world_model,0.745319793621699,0.7152381887038549,5.0,0.9
+unitree_z1_stackbox,case2,sparse_8,policy,1.0000000099341075,1.000000034769376,8.0,0.84
+unitree_z1_stackbox,case2,sparse_8,world_model,1.0000000198682149,1.0000000149011612,8.0,0.84
+unitree_z1_stackbox,case2,tail_heavy_10,policy,1.0000000099341075,1.000000034769376,10.0,0.8000000000000002
+unitree_z1_stackbox,case2,tail_heavy_10,world_model,1.0000000198682149,1.0000000149011612,10.0,0.8000000000000002
+unitree_z1_stackbox,case2,tail_heavy_6,policy,1.0000000099341075,1.000000034769376,6.0,0.88
+unitree_z1_stackbox,case2,tail_heavy_6,world_model,1.0000000198682149,1.0000000149011612,6.0,0.88
+unitree_z1_stackbox,case2,tail_heavy_8,policy,1.0000000099341075,1.000000034769376,8.0,0.84
+unitree_z1_stackbox,case2,tail_heavy_8,world_model,1.0000000198682149,1.0000000149011612,8.0,0.84
+unitree_z1_stackbox,case2,tail_only_4,policy,1.0000000099341075,1.000000034769376,4.0,0.92
+unitree_z1_stackbox,case2,tail_only_4,world_model,1.0000000198682149,1.0000000149011612,4.0,0.92
+unitree_z1_stackbox,case2,tail_only_6,policy,1.0000000099341075,1.000000034769376,6.0,0.88
+unitree_z1_stackbox,case2,tail_only_6,world_model,1.0000000198682149,1.0000000149011612,6.0,0.88
+unitree_z1_stackbox,case3,dense,policy,1.0000000447034836,0.9999999950329462,50.0,0.0
+unitree_z1_stackbox,case3,dense,world_model,1.0000000298023224,1.0,50.0,0.0
+unitree_z1_stackbox,case3,sparse_10,policy,0.9272978107134501,0.8970760951439539,10.0,0.8000000000000002
+unitree_z1_stackbox,case3,sparse_10,world_model,0.9226573010285696,0.8924628843863806,10.0,0.8000000000000002
+unitree_z1_stackbox,case3,sparse_4,policy,1.0000000447034836,0.9999999950329462,4.0,0.92
+unitree_z1_stackbox,case3,sparse_4,world_model,1.0000000298023224,1.0,4.0,0.92
+unitree_z1_stackbox,case3,sparse_5,policy,0.7838167995214462,0.7392430454492569,5.0,0.9
+unitree_z1_stackbox,case3,sparse_5,world_model,0.7687370975812277,0.7268570611874262,5.0,0.9
+unitree_z1_stackbox,case3,sparse_8,policy,1.0000000447034836,0.9999999950329462,8.0,0.84
+unitree_z1_stackbox,case3,sparse_8,world_model,1.0000000298023224,1.0,8.0,0.84
+unitree_z1_stackbox,case3,tail_heavy_10,policy,1.0000000447034836,0.9999999950329462,10.0,0.8000000000000002
+unitree_z1_stackbox,case3,tail_heavy_10,world_model,1.0000000298023224,1.0,10.0,0.8000000000000002
+unitree_z1_stackbox,case3,tail_heavy_6,policy,1.0000000447034836,0.9999999950329462,6.0,0.88
+unitree_z1_stackbox,case3,tail_heavy_6,world_model,1.0000000298023224,1.0,6.0,0.88
+unitree_z1_stackbox,case3,tail_heavy_8,policy,1.0000000447034836,0.9999999950329462,8.0,0.84
+unitree_z1_stackbox,case3,tail_heavy_8,world_model,1.0000000298023224,1.0,8.0,0.84
+unitree_z1_stackbox,case3,tail_only_4,policy,1.0000000447034836,0.9999999950329462,4.0,0.92
+unitree_z1_stackbox,case3,tail_only_4,world_model,1.0000000298023224,1.0,4.0,0.92
+unitree_z1_stackbox,case3,tail_only_6,policy,1.0000000447034836,0.9999999950329462,6.0,0.88
+unitree_z1_stackbox,case3,tail_only_6,world_model,1.0000000298023224,1.0,6.0,0.88
+unitree_z1_stackbox,case4,dense,policy,1.00000003973643,1.0000000298023224,50.0,0.0
+unitree_z1_stackbox,case4,dense,world_model,1.0000000099341075,1.0000000248352687,50.0,0.0
+unitree_z1_stackbox,case4,sparse_10,policy,0.9115401953458786,0.8695555130640665,10.0,0.8000000000000002
+unitree_z1_stackbox,case4,sparse_10,world_model,0.9073014905055364,0.8689398914575577,10.0,0.8000000000000002
+unitree_z1_stackbox,case4,sparse_4,policy,1.00000003973643,1.0000000298023224,4.0,0.92
+unitree_z1_stackbox,case4,sparse_4,world_model,1.0000000099341075,1.0000000248352687,4.0,0.92
+unitree_z1_stackbox,case4,sparse_5,policy,0.7569213857253393,0.6986446877320608,5.0,0.9
+unitree_z1_stackbox,case4,sparse_5,world_model,0.7459165553251902,0.6989829589923223,5.0,0.9
+unitree_z1_stackbox,case4,sparse_8,policy,1.00000003973643,1.0000000298023224,8.0,0.84
+unitree_z1_stackbox,case4,sparse_8,world_model,1.0000000099341075,1.0000000248352687,8.0,0.84
+unitree_z1_stackbox,case4,tail_heavy_10,policy,1.00000003973643,1.0000000298023224,10.0,0.8000000000000002
+unitree_z1_stackbox,case4,tail_heavy_10,world_model,1.0000000099341075,1.0000000248352687,10.0,0.8000000000000002
+unitree_z1_stackbox,case4,tail_heavy_6,policy,1.00000003973643,1.0000000298023224,6.0,0.88
+unitree_z1_stackbox,case4,tail_heavy_6,world_model,1.0000000099341075,1.0000000248352687,6.0,0.88
+unitree_z1_stackbox,case4,tail_heavy_8,policy,1.00000003973643,1.0000000298023224,8.0,0.84
+unitree_z1_stackbox,case4,tail_heavy_8,world_model,1.0000000099341075,1.0000000248352687,8.0,0.84
+unitree_z1_stackbox,case4,tail_only_4,policy,1.00000003973643,1.0000000298023224,4.0,0.92
+unitree_z1_stackbox,case4,tail_only_4,world_model,1.0000000099341075,1.0000000248352687,4.0,0.92
+unitree_z1_stackbox,case4,tail_only_6,policy,1.00000003973643,1.0000000298023224,6.0,0.88
+unitree_z1_stackbox,case4,tail_only_6,world_model,1.0000000099341075,1.0000000248352687,6.0,0.88

src/unifolm_wma/models/samplers/ddim.py

@@ -6,6 +6,8 @@ from unifolm_wma.utils.diffusion import make_ddim_sampling_parameters, make_ddim
 from unifolm_wma.utils.common import noise_like
 from unifolm_wma.utils.common import extract_into_tensor
 from tqdm import tqdm
+from unifolm_wma.modules.attention import enable_cross_attn_kv_cache, disable_cross_attn_kv_cache
+from unifolm_wma.modules.networks.wma_model import enable_ctx_cache, disable_ctx_cache
 
 
 class DDIMSampler(object):
@@ -19,8 +21,9 @@ class DDIMSampler(object):
 
     def register_buffer(self, name, attr):
         if type(attr) == torch.Tensor:
-            if attr.device != torch.device("cuda"):
+            target_device = self.model.device
-                attr = attr.to(torch.device("cuda"))
+            if attr.device != target_device:
+                attr = attr.to(target_device)
         setattr(self, name, attr)
 
     def make_schedule(self,
@@ -67,11 +70,12 @@ class DDIMSampler(object):
             ddim_timesteps=self.ddim_timesteps,
             eta=ddim_eta,
             verbose=verbose)
-        self.register_buffer('ddim_sigmas', ddim_sigmas)
+        # Ensure tensors are on correct device for efficient indexing
-        self.register_buffer('ddim_alphas', ddim_alphas)
+        self.register_buffer('ddim_sigmas', to_torch(torch.as_tensor(ddim_sigmas)))
-        self.register_buffer('ddim_alphas_prev', ddim_alphas_prev)
+        self.register_buffer('ddim_alphas', to_torch(torch.as_tensor(ddim_alphas)))
+        self.register_buffer('ddim_alphas_prev', to_torch(torch.as_tensor(ddim_alphas_prev)))
         self.register_buffer('ddim_sqrt_one_minus_alphas',
-                             np.sqrt(1. - ddim_alphas))
+                             to_torch(torch.as_tensor(np.sqrt(1. - ddim_alphas))))
         sigmas_for_original_sampling_steps = ddim_eta * torch.sqrt(
             (1 - self.alphas_cumprod_prev) / (1 - self.alphas_cumprod) *
             (1 - self.alphas_cumprod / self.alphas_cumprod_prev))
@@ -106,6 +110,9 @@ class DDIMSampler(object):
             fs=None,
             timestep_spacing='uniform',  #uniform_trailing for starting from last timestep
             guidance_rescale=0.0,
+            handoff_step: int = 0,
+            handoff_callback=None,
+            stop_at_handoff: bool = False,
             **kwargs):
 
         # Check condition bs
@@ -161,6 +168,9 @@ class DDIMSampler(object):
             precision=precision,
             fs=fs,
             guidance_rescale=guidance_rescale,
+            handoff_step=handoff_step,
+            handoff_callback=handoff_callback,
+            stop_at_handoff=stop_at_handoff,
             **kwargs)
         return samples, actions, states, intermediates
 
@@ -187,6 +197,9 @@ class DDIMSampler(object):
                       precision=None,
                       fs=None,
                       guidance_rescale=0.0,
+                      handoff_step: int = 0,
+                      handoff_callback=None,
+                      stop_at_handoff: bool = False,
                       **kwargs):
         device = self.model.betas.device
         dp_ddim_scheduler_action = self.model.dp_noise_scheduler_action
@@ -195,22 +208,12 @@ class DDIMSampler(object):
         b = shape[0]
         if x_T is None:
             img = torch.randn(shape, device=device)
-            action = torch.randn((b, 16, self.model.agent_action_dim),
+            action = torch.randn((b, 16, self.model.agent_action_dim), device=device)
-                                 device=device)
+            state = torch.randn((b, 16, self.model.agent_state_dim), device=device)
-            state = torch.randn((b, 16, self.model.agent_state_dim),
-                                device=device)
         else:
             img = x_T
-            action = torch.randn((b, 16, self.model.agent_action_dim),
+            action = torch.randn((b, 16, self.model.agent_action_dim), device=device)
-                                 device=device)
+            state = torch.randn((b, 16, self.model.agent_state_dim), device=device)
-            state = torch.randn((b, 16, self.model.agent_state_dim),
-                                device=device)
-
-        if precision is not None:
-            if precision == 16:
-                img = img.to(dtype=torch.float16)
-                action = action.to(dtype=torch.float16)
-                state = state.to(dtype=torch.float16)
 
         if timesteps is None:
             timesteps = self.ddpm_num_timesteps if ddim_use_original_steps else self.ddim_timesteps
@@ -238,67 +241,91 @@ class DDIMSampler(object):
             iterator = time_range
 
         clean_cond = kwargs.pop("clean_cond", False)
-
         dp_ddim_scheduler_action.set_timesteps(len(timesteps))
         dp_ddim_scheduler_state.set_timesteps(len(timesteps))
-        for i, step in enumerate(iterator):
+        ts = torch.empty((b, ), device=device, dtype=torch.long)
-            index = total_steps - i - 1
+        enable_cross_attn_kv_cache(self.model)
-            ts = torch.full((b, ), step, device=device, dtype=torch.long)
+        enable_ctx_cache(self.model)
+        handoff = {}
+        try:
+            for i, step in enumerate(iterator):
+                index = total_steps - i - 1
+                ts.fill_(step)
 
-            # Use mask to blend noised original latent (img_orig) & new sampled latent (img)
+                # Use mask to blend noised original latent (img_orig) & new sampled latent (img)
-            if mask is not None:
+                if mask is not None:
-                assert x0 is not None
+                    assert x0 is not None
-                if clean_cond:
+                    if clean_cond:
-                    img_orig = x0
+                        img_orig = x0
-                else:
+                    else:
-                    img_orig = self.model.q_sample(x0, ts)
+                        img_orig = self.model.q_sample(x0, ts)
-                img = img_orig * mask + (1. - mask) * img
+                    img = img_orig * mask + (1. - mask) * img
 
-            outs = self.p_sample_ddim(
+                outs = self.p_sample_ddim(
-                img,
+                    img,
-                action,
+                    action,
-                state,
+                    state,
-                cond,
+                    cond,
-                ts,
+                    ts,
-                index=index,
+                    index=index,
-                use_original_steps=ddim_use_original_steps,
+                    precision=precision,
-                quantize_denoised=quantize_denoised,
+                    use_original_steps=ddim_use_original_steps,
-                temperature=temperature,
+                    quantize_denoised=quantize_denoised,
-                noise_dropout=noise_dropout,
+                    temperature=temperature,
-                score_corrector=score_corrector,
+                    noise_dropout=noise_dropout,
-                corrector_kwargs=corrector_kwargs,
+                    score_corrector=score_corrector,
-                unconditional_guidance_scale=unconditional_guidance_scale,
+                    corrector_kwargs=corrector_kwargs,
-                unconditional_conditioning=unconditional_conditioning,
+                    unconditional_guidance_scale=unconditional_guidance_scale,
-                mask=mask,
+                    unconditional_conditioning=unconditional_conditioning,
-                x0=x0,
+                    mask=mask,
-                fs=fs,
+                    x0=x0,
-                guidance_rescale=guidance_rescale,
+                    fs=fs,
-                **kwargs)
+                    guidance_rescale=guidance_rescale,
+                    **kwargs)
 
-            img, pred_x0, model_output_action, model_output_state = outs
+                img, pred_x0, model_output_action, model_output_state = outs
 
-            action = dp_ddim_scheduler_action.step(
+                action = dp_ddim_scheduler_action.step(
-                model_output_action,
+                    model_output_action,
-                step,
+                    step,
-                action,
+                    action,
-                generator=None,
+                    generator=None,
-            ).prev_sample
+                ).prev_sample
-            state = dp_ddim_scheduler_state.step(
+                state = dp_ddim_scheduler_state.step(
-                model_output_state,
+                    model_output_state,
-                step,
+                    step,
-                state,
+                    state,
-                generator=None,
+                    generator=None,
-            ).prev_sample
+                ).prev_sample
 
-            if callback: callback(i)
+                if callback: callback(i)
-            if img_callback: img_callback(pred_x0, i)
+                if img_callback: img_callback(pred_x0, i)
 
-            if index % log_every_t == 0 or index == total_steps - 1:
+                if handoff_step > 0 and (i + 1) == handoff_step:
-                intermediates['x_inter'].append(img)
+                    handoff = {
-                intermediates['pred_x0'].append(pred_x0)
+                        'samples': img.clone(),
-                intermediates['x_inter_action'].append(action)
+                        'actions': action.clone(),
-                intermediates['x_inter_state'].append(state)
+                        'states': state.clone(),
+                        'pred_x0': pred_x0.clone(),
+                        'step': i + 1,
+                    }
+                    if handoff_callback is not None:
+                        handoff_callback(handoff)
+                    if stop_at_handoff:
+                        intermediates['handoff'] = handoff
+                        break
 
+                if index % log_every_t == 0 or index == total_steps - 1:
+                    intermediates['x_inter'].append(img)
+                    intermediates['pred_x0'].append(pred_x0)
+                    intermediates['x_inter_action'].append(action)
+                    intermediates['x_inter_state'].append(state)
+        finally:
+            disable_cross_attn_kv_cache(self.model)
+            disable_ctx_cache(self.model)
+
+        if handoff_step > 0:
+            intermediates['handoff'] = handoff
         return img, action, state, intermediates
 
     @torch.no_grad()
@@ -309,6 +336,7 @@ class DDIMSampler(object):
                       c,
                       t,
                       index,
+                      precision=None,
                       repeat_noise=False,
                       use_original_steps=False,
                       quantize_denoised=False,
@@ -325,42 +353,41 @@ class DDIMSampler(object):
                       guidance_rescale=0.0,
                       **kwargs):
         b, *_, device = *x.shape, x.device
-        if x.dim() == 5:
-            is_video = True
-        else:
-            is_video = False
 
-        if unconditional_conditioning is None or unconditional_guidance_scale == 1.:
+        use_autocast = precision == 16 and device.type == 'cuda'
-            model_output, model_output_action, model_output_state = self.model.apply_model(
+        with torch.cuda.amp.autocast(enabled=use_autocast,
-                x, x_action, x_state, t, c, **kwargs)  # unet denoiser
+                                     dtype=torch.float16):
-        else:
+            if unconditional_conditioning is None or unconditional_guidance_scale == 1.:
-            # do_classifier_free_guidance
+                model_output, model_output_action, model_output_state = self.model.apply_model(
-            if isinstance(c, torch.Tensor) or isinstance(c, dict):
+                    x, x_action, x_state, t, c, **kwargs)  # unet denoiser
-                e_t_cond, e_t_cond_action, e_t_cond_state = self.model.apply_model(
-                    x, x_action, x_state, t, c, **kwargs)
-                e_t_uncond, e_t_uncond_action, e_t_uncond_state = self.model.apply_model(
-                    x, x_action, x_state, t, unconditional_conditioning,
-                    **kwargs)
             else:
-                raise NotImplementedError
+                # do_classifier_free_guidance
-            model_output = e_t_uncond + unconditional_guidance_scale * (
+                if isinstance(c, torch.Tensor) or isinstance(c, dict):
-                e_t_cond - e_t_uncond)
+                    e_t_cond, e_t_cond_action, e_t_cond_state = self.model.apply_model(
-            model_output_action = e_t_uncond_action + unconditional_guidance_scale * (
+                        x, x_action, x_state, t, c, **kwargs)
-                e_t_cond_action - e_t_uncond_action)
+                    e_t_uncond, e_t_uncond_action, e_t_uncond_state = self.model.apply_model(
-            model_output_state = e_t_uncond_state + unconditional_guidance_scale * (
+                        x, x_action, x_state, t, unconditional_conditioning,
-                e_t_cond_state - e_t_uncond_state)
+                        **kwargs)
+                else:
+                    raise NotImplementedError
+                model_output = e_t_uncond + unconditional_guidance_scale * (
+                    e_t_cond - e_t_uncond)
+                model_output_action = e_t_uncond_action + unconditional_guidance_scale * (
+                    e_t_cond_action - e_t_uncond_action)
+                model_output_state = e_t_uncond_state + unconditional_guidance_scale * (
+                    e_t_cond_state - e_t_uncond_state)
 
-            if guidance_rescale > 0.0:
+                if guidance_rescale > 0.0:
-                model_output = rescale_noise_cfg(
+                    model_output = rescale_noise_cfg(
-                    model_output, e_t_cond, guidance_rescale=guidance_rescale)
+                        model_output, e_t_cond, guidance_rescale=guidance_rescale)
-                model_output_action = rescale_noise_cfg(
+                    model_output_action = rescale_noise_cfg(
-                    model_output_action,
+                        model_output_action,
-                    e_t_cond_action,
+                        e_t_cond_action,
-                    guidance_rescale=guidance_rescale)
+                        guidance_rescale=guidance_rescale)
-                model_output_state = rescale_noise_cfg(
+                    model_output_state = rescale_noise_cfg(
-                    model_output_state,
+                        model_output_state,
-                    e_t_cond_state,
+                        e_t_cond_state,
-                    guidance_rescale=guidance_rescale)
+                        guidance_rescale=guidance_rescale)
 
         if self.model.parameterization == "v":
             e_t = self.model.predict_eps_from_z_and_v(x, t, model_output)
@@ -377,17 +404,11 @@ class DDIMSampler(object):
         sqrt_one_minus_alphas = self.model.sqrt_one_minus_alphas_cumprod if use_original_steps else self.ddim_sqrt_one_minus_alphas
         sigmas = self.ddim_sigmas_for_original_num_steps if use_original_steps else self.ddim_sigmas
 
-        if is_video:
+        # Use 0-d tensors directly (already on device); broadcasting handles shape
-            size = (b, 1, 1, 1, 1)
+        a_t = alphas[index]
-        else:
+        a_prev = alphas_prev[index]
-            size = (b, 1, 1, 1)
+        sigma_t = sigmas[index]
-
+        sqrt_one_minus_at = sqrt_one_minus_alphas[index]
-        a_t = torch.full(size, alphas[index], device=device)
-        a_prev = torch.full(size, alphas_prev[index], device=device)
-        sigma_t = torch.full(size, sigmas[index], device=device)
-        sqrt_one_minus_at = torch.full(size,
-                                       sqrt_one_minus_alphas[index],
-                                       device=device)
 
         if self.model.parameterization != "v":
             pred_x0 = (x - sqrt_one_minus_at * e_t) / a_t.sqrt()
@@ -395,12 +416,8 @@ class DDIMSampler(object):
             pred_x0 = self.model.predict_start_from_z_and_v(x, t, model_output)
 
         if self.model.use_dynamic_rescale:
-            scale_t = torch.full(size,
+            scale_t = self.ddim_scale_arr[index]
-                                 self.ddim_scale_arr[index],
+            prev_scale_t = self.ddim_scale_arr_prev[index]
-                                 device=device)
-            prev_scale_t = torch.full(size,
-                                      self.ddim_scale_arr_prev[index],
-                                      device=device)
             rescale = (prev_scale_t / scale_t)
             pred_x0 *= rescale
 

src/unifolm_wma/modules/networks/wma_model.py

@@ -685,6 +685,37 @@ class WMAModel(nn.Module):
         self.action_token_projector = instantiate_from_config(
             stem_process_config)
 
+        # Context precomputation cache
+        self._ctx_cache_enabled = False
+        self._ctx_cache = {}
+        self._trt_backbone = None  # TRT engine for video UNet backbone
+        # Reusable CUDA stream for parallel state_unet / action_unet
+        self._state_stream = torch.cuda.Stream()
+
+    def __getstate__(self):
+        state = self.__dict__.copy()
+        state.pop('_state_stream', None)
+        return state
+
+    def __setstate__(self, state):
+        self.__dict__.update(state)
+        if not hasattr(self, '_ctx_cache_enabled'):
+            self._ctx_cache_enabled = False
+        if not hasattr(self, '_ctx_cache'):
+            self._ctx_cache = {}
+        if not hasattr(self, '_trt_backbone'):
+            self._trt_backbone = None
+        self._state_stream = torch.cuda.Stream()
+
+    def load_trt_backbone(self, engine_path, n_hs_a=9):
+        """Load a TensorRT engine for the video UNet backbone."""
+        from unifolm_wma.trt_utils import TRTBackbone
+        device = next(self.parameters()).device
+        self._trt_backbone = TRTBackbone(engine_path,
+                                         n_hs_a=n_hs_a,
+                                         device=device)
+        print(f">>> TRT backbone loaded from {engine_path} on {device}")
+
     def forward(self,
                 x: Tensor,
                 x_action: Tensor,
@@ -713,65 +744,70 @@ class WMAModel(nn.Module):
             Tuple of Tensors for predictions:
 
         """
-
         b, _, t, _, _ = x.shape
         t_emb = timestep_embedding(timesteps,
                                    self.model_channels,
                                    repeat_only=False).type(x.dtype)
         emb = self.time_embed(t_emb)
 
-        bt, l_context, _ = context.shape
+        _ctx_key = context.data_ptr()
-        if self.base_model_gen_only:
+        if self._ctx_cache_enabled and _ctx_key in self._ctx_cache:
-            assert l_context == 77 + self.n_obs_steps * 16, ">>> ERROR Context dim 1 ..."  ## NOTE HANDCODE
+            context = self._ctx_cache[_ctx_key]
         else:
-            if l_context == self.n_obs_steps + 77 + t * 16:
+            bt, l_context, _ = context.shape
-                context_agent_state = context[:, :self.n_obs_steps]
+            if self.base_model_gen_only:
-                context_text = context[:, self.n_obs_steps:self.n_obs_steps +
+                assert l_context == 77 + self.n_obs_steps * 16, ">>> ERROR Context dim 1 ..."  ## NOTE HANDCODE
-                                       77, :]
+            else:
-                context_img = context[:, self.n_obs_steps + 77:, :]
+                if l_context == self.n_obs_steps + 77 + t * 16:
-                context_agent_state = context_agent_state.repeat_interleave(
+                    context_agent_state = context[:, :self.n_obs_steps]
-                    repeats=t, dim=0)
+                    context_text = context[:, self.n_obs_steps:self.n_obs_steps +
-                context_text = context_text.repeat_interleave(repeats=t, dim=0)
+                                           77, :]
-                context_img = rearrange(context_img,
+                    context_img = context[:, self.n_obs_steps + 77:, :]
-                                        'b (t l) c -> (b t) l c',
+                    context_agent_state = context_agent_state.repeat_interleave(
-                                        t=t)
+                        repeats=t, dim=0)
-                context = torch.cat(
+                    context_text = context_text.repeat_interleave(repeats=t, dim=0)
-                    [context_agent_state, context_text, context_img], dim=1)
+                    context_img = rearrange(context_img,
-            elif l_context == self.n_obs_steps + 16 + 77 + t * 16:
+                                            'b (t l) c -> (b t) l c',
-                context_agent_state = context[:, :self.n_obs_steps]
+                                            t=t)
-                context_agent_action = context[:, self.
+                    context = torch.cat(
-                                               n_obs_steps:self.n_obs_steps +
+                        [context_agent_state, context_text, context_img], dim=1)
-                                               16, :]
+                elif l_context == self.n_obs_steps + 16 + 77 + t * 16:
-                context_agent_action = rearrange(
+                    context_agent_state = context[:, :self.n_obs_steps]
-                    context_agent_action.unsqueeze(2), 'b t l d -> (b t) l d')
+                    context_agent_action = context[:, self.
-                context_agent_action = self.action_token_projector(
+                                                   n_obs_steps:self.n_obs_steps +
-                    context_agent_action)
+                                                   16, :]
-                context_agent_action = rearrange(context_agent_action,
+                    context_agent_action = rearrange(
-                                                 '(b o) l d -> b o l d',
+                        context_agent_action.unsqueeze(2), 'b t l d -> (b t) l d')
-                                                 o=t)
+                    context_agent_action = self.action_token_projector(
-                context_agent_action = rearrange(context_agent_action,
+                        context_agent_action)
-                                                 'b o (t l) d -> b o t l d',
+                    context_agent_action = rearrange(context_agent_action,
-                                                 t=t)
+                                                     '(b o) l d -> b o l d',
-                context_agent_action = context_agent_action.permute(
+                                                     o=t)
-                    0, 2, 1, 3, 4)
+                    context_agent_action = rearrange(context_agent_action,
-                context_agent_action = rearrange(context_agent_action,
+                                                     'b o (t l) d -> b o t l d',
-                                                 'b t o l d -> (b t) (o l) d')
+                                                     t=t)
+                    context_agent_action = context_agent_action.permute(
+                        0, 2, 1, 3, 4)
+                    context_agent_action = rearrange(context_agent_action,
+                                                     'b t o l d -> (b t) (o l) d')
 
-                context_text = context[:, self.n_obs_steps +
+                    context_text = context[:, self.n_obs_steps +
-                                       16:self.n_obs_steps + 16 + 77, :]
+                                           16:self.n_obs_steps + 16 + 77, :]
-                context_text = context_text.repeat_interleave(repeats=t, dim=0)
+                    context_text = context_text.repeat_interleave(repeats=t, dim=0)
 
-                context_img = context[:, self.n_obs_steps + 16 + 77:, :]
+                    context_img = context[:, self.n_obs_steps + 16 + 77:, :]
-                context_img = rearrange(context_img,
+                    context_img = rearrange(context_img,
-                                        'b (t l) c -> (b t) l c',
+                                            'b (t l) c -> (b t) l c',
-                                        t=t)
+                                            t=t)
-                context_agent_state = context_agent_state.repeat_interleave(
+                    context_agent_state = context_agent_state.repeat_interleave(
-                    repeats=t, dim=0)
+                        repeats=t, dim=0)
-                context = torch.cat([
+                    context = torch.cat([
-                    context_agent_state, context_agent_action, context_text,
+                        context_agent_state, context_agent_action, context_text,
-                    context_img
+                        context_img
-                ],
+                    ],
-                                    dim=1)
+                                        dim=1)
+            if self._ctx_cache_enabled:
+                self._ctx_cache[_ctx_key] = context
 
         emb = emb.repeat_interleave(repeats=t, dim=0)
 
@@ -791,58 +827,95 @@ class WMAModel(nn.Module):
             fs_embed = fs_embed.repeat_interleave(repeats=t, dim=0)
             emb = emb + fs_embed
 
-        h = x.type(self.dtype)
+        if self._trt_backbone is not None:
-        adapter_idx = 0
+            # TRT path: run backbone via TensorRT engine
-        hs = []
+            h_in = x.type(self.dtype).contiguous()
-        hs_a = []
+            y, hs_a = self._trt_backbone(h_in, emb.contiguous(), context.contiguous())
-        for id, module in enumerate(self.input_blocks):
+        else:
-            h = module(h, emb, context=context, batch_size=b)
+            # PyTorch path: original backbone
-            if id == 0 and self.addition_attention:
+            h = x.type(self.dtype)
-                h = self.init_attn(h, emb, context=context, batch_size=b)
+            adapter_idx = 0
-            # plug-in adapter features
+            hs = []
-            if ((id + 1) % 3 == 0) and features_adapter is not None:
+            hs_a = []
-                h = h + features_adapter[adapter_idx]
+            for id, module in enumerate(self.input_blocks):
-                adapter_idx += 1
+                h = module(h, emb, context=context, batch_size=b)
-            if id != 0:
+                if id == 0 and self.addition_attention:
-                if isinstance(module[0], Downsample):
+                    h = self.init_attn(h, emb, context=context, batch_size=b)
+                # plug-in adapter features
+                if ((id + 1) % 3 == 0) and features_adapter is not None:
+                    h = h + features_adapter[adapter_idx]
+                    adapter_idx += 1
+                if id != 0:
+                    if isinstance(module[0], Downsample):
+                        hs_a.append(
+                            rearrange(hs[-1], '(b t) c h w -> b t c h w', b=b))
+                hs.append(h)
+            hs_a.append(rearrange(h, '(b t) c h w -> b t c h w', b=b))
+
+            if features_adapter is not None:
+                assert len(
+                    features_adapter) == adapter_idx, 'Wrong features_adapter'
+            h = self.middle_block(h, emb, context=context, batch_size=b)
+            hs_a.append(rearrange(h, '(b t) c h w -> b t c h w', b=b))
+
+            hs_out = []
+            for module in self.output_blocks:
+                h = torch.cat([h, hs.pop()], dim=1)
+                h = module(h, emb, context=context, batch_size=b)
+                if isinstance(module[-1], Upsample):
                     hs_a.append(
-                        rearrange(hs[-1], '(b t) c h w -> b t c h w', t=t))
+                        rearrange(hs_out[-1], '(b t) c h w -> b t c h w', b=b))
-            hs.append(h)
+                hs_out.append(h)
-        hs_a.append(rearrange(h, '(b t) c h w -> b t c h w', t=t))
+            h = h.type(x.dtype)
+            hs_a.append(rearrange(hs_out[-1], '(b t) c h w -> b t c h w', b=b))
 
-        if features_adapter is not None:
+            y = self.out(h)
-            assert len(
+            y = rearrange(y, '(b t) c h w -> b c t h w', b=b)
-                features_adapter) == adapter_idx, 'Wrong features_adapter'
-        h = self.middle_block(h, emb, context=context, batch_size=b)
-        hs_a.append(rearrange(h, '(b t) c h w -> b t c h w', t=t))
-
-        hs_out = []
-        for module in self.output_blocks:
-            h = torch.cat([h, hs.pop()], dim=1)
-            h = module(h, emb, context=context, batch_size=b)
-            if isinstance(module[-1], Upsample):
-                hs_a.append(
-                    rearrange(hs_out[-1], '(b t) c h w -> b t c h w', t=t))
-            hs_out.append(h)
-        h = h.type(x.dtype)
-        hs_a.append(rearrange(hs_out[-1], '(b t) c h w -> b t c h w', t=t))
-
-        y = self.out(h)
-        y = rearrange(y, '(b t) c h w -> b c t h w', b=b)
 
         if not self.base_model_gen_only:
             ba, _, _ = x_action.shape
-            a_y = self.action_unet(x_action, timesteps[:ba], hs_a,
+            ts_state = timesteps[:ba] if b > 1 else timesteps
-                                   context_action[:2], **kwargs)
+            is_sim_mode = context_action[2] if len(context_action) > 2 else False
-            # Predict state
+
-            if b > 1:
+            if is_sim_mode:
-                s_y = self.state_unet(x_state, timesteps[:ba], hs_a,
+                # WM mode: only need state_unet, skip action_unet
+                s_y = self.state_unet(x_state, ts_state, hs_a,
                                       context_action[:2], **kwargs)
+                a_y = torch.zeros_like(x_action)
             else:
-                s_y = self.state_unet(x_state, timesteps, hs_a,
+                # DM mode: only need action_unet, skip state_unet
-                                      context_action[:2], **kwargs)
+                a_y = self.action_unet(x_action, timesteps[:ba], hs_a,
+                                       context_action[:2], **kwargs)
+                s_y = torch.zeros_like(x_state)
         else:
             a_y = torch.zeros_like(x_action)
             s_y = torch.zeros_like(x_state)
 
         return y, a_y, s_y
+
+
+def enable_ctx_cache(model):
+    """Enable context precomputation cache on WMAModel and its action/state UNets."""
+    for m in model.modules():
+        if isinstance(m, WMAModel):
+            m._ctx_cache_enabled = True
+            m._ctx_cache = {}
+    # conditional_unet1d cache
+    from unifolm_wma.models.diffusion_head.conditional_unet1d import ConditionalUnet1D
+    for m in model.modules():
+        if isinstance(m, ConditionalUnet1D):
+            m._global_cond_cache_enabled = True
+            m._global_cond_cache = {}
+
+
+def disable_ctx_cache(model):
+    """Disable and clear context precomputation cache."""
+    for m in model.modules():
+        if isinstance(m, WMAModel):
+            m._ctx_cache_enabled = False
+            m._ctx_cache = {}
+    from unifolm_wma.models.diffusion_head.conditional_unet1d import ConditionalUnet1D
+    for m in model.modules():
+        if isinstance(m, ConditionalUnet1D):
+            m._global_cond_cache_enabled = False
+            m._global_cond_cache = {}

src/unifolm_wma/trt_utils.py

@@ -0,0 +1,196 @@
+"""TensorRT acceleration utilities for the video UNet backbone."""
+
+import torch
+import torch.nn as nn
+from einops import rearrange
+from unifolm_wma.modules.networks.wma_model import Downsample, Upsample
+
+
+def _normalize_cuda_device(device) -> torch.device:
+    if device is None:
+        return torch.device('cuda', torch.cuda.current_device())
+    if isinstance(device, torch.device):
+        if device.type != 'cuda':
+            raise ValueError(f"TensorRT requires a CUDA device, got {device}.")
+        return device
+    if isinstance(device, int):
+        return torch.device('cuda', device)
+    normalized = torch.device(device)
+    if normalized.type != 'cuda':
+        raise ValueError(f"TensorRT requires a CUDA device, got {normalized}.")
+    return normalized
+
+
+class VideoBackboneForExport(nn.Module):
+    """Wrapper that isolates the video UNet backbone for ONNX export.
+
+    Takes already-preprocessed inputs (after context/time embedding prep)
+    and returns y + hs_a as a flat tuple.
+    """
+
+    def __init__(self, wma_model):
+        super().__init__()
+        self.input_blocks = wma_model.input_blocks
+        self.middle_block = wma_model.middle_block
+        self.output_blocks = wma_model.output_blocks
+        self.out = wma_model.out
+        self.addition_attention = wma_model.addition_attention
+        if self.addition_attention:
+            self.init_attn = wma_model.init_attn
+        self.dtype = wma_model.dtype
+
+    def forward(self, h, emb, context):
+        t = 16
+        b = 1
+
+        hs = []
+        hs_a = []
+        h = h.type(self.dtype)
+        for id, module in enumerate(self.input_blocks):
+            h = module(h, emb, context=context, batch_size=b)
+            if id == 0 and self.addition_attention:
+                h = self.init_attn(h, emb, context=context, batch_size=b)
+            if id != 0:
+                if isinstance(module[0], Downsample):
+                    hs_a.append(rearrange(hs[-1], '(b t) c h w -> b t c h w', t=t))
+            hs.append(h)
+        hs_a.append(rearrange(h, '(b t) c h w -> b t c h w', t=t))
+
+        h = self.middle_block(h, emb, context=context, batch_size=b)
+        hs_a.append(rearrange(h, '(b t) c h w -> b t c h w', t=t))
+
+        hs_out = []
+        for module in self.output_blocks:
+            h = torch.cat([h, hs.pop()], dim=1)
+            h = module(h, emb, context=context, batch_size=b)
+            if isinstance(module[-1], Upsample):
+                hs_a.append(rearrange(hs_out[-1], '(b t) c h w -> b t c h w', t=t))
+            hs_out.append(h)
+        hs_a.append(rearrange(hs_out[-1], '(b t) c h w -> b t c h w', t=t))
+
+        y = self.out(h.type(h.dtype))
+        y = rearrange(y, '(b t) c h w -> b c t h w', b=b)
+        return (y, *hs_a)
+
+
+def export_backbone_onnx(model, save_path, context_len=95):
+    wma = model.model.diffusion_model
+    wrapper = VideoBackboneForExport(wma)
+    device = next(wma.parameters()).device
+    wrapper.eval().to(device)
+
+    for m in wrapper.modules():
+        if hasattr(m, 'checkpoint'):
+            m.checkpoint = False
+        if hasattr(m, 'use_checkpoint'):
+            m.use_checkpoint = False
+
+    import xformers.ops
+    _orig_mea = xformers.ops.memory_efficient_attention
+    def _sdpa_replacement(q, k, v, attn_bias=None, op=None, **kw):
+        return torch.nn.functional.scaled_dot_product_attention(q, k, v, attn_mask=attn_bias)
+    xformers.ops.memory_efficient_attention = _sdpa_replacement
+
+    BT = 16
+    emb_dim = wma.model_channels * 4
+    ctx_dim = 1024
+    in_ch = wma.in_channels
+
+    dummy_h = torch.randn(BT,
+                          in_ch,
+                          40,
+                          64,
+                          device=device,
+                          dtype=torch.float32)
+    dummy_emb = torch.randn(BT,
+                            emb_dim,
+                            device=device,
+                            dtype=torch.float32)
+    dummy_ctx = torch.randn(BT,
+                            context_len,
+                            ctx_dim,
+                            device=device,
+                            dtype=torch.float32)
+
+    with torch.no_grad():
+        outputs = wrapper(dummy_h, dummy_emb, dummy_ctx)
+        n_outputs = len(outputs)
+        print(f">>> Backbone has {n_outputs} outputs (1 y + {n_outputs-1} hs_a)")
+        for i, o in enumerate(outputs):
+            print(f"    output[{i}]: {o.shape} {o.dtype}")
+
+    output_names = ['y'] + [f'hs_a_{i}' for i in range(n_outputs - 1)]
+
+    torch.onnx.export(
+        wrapper,
+        (dummy_h, dummy_emb, dummy_ctx),
+        save_path,
+        input_names=['h', 'emb', 'context'],
+        output_names=output_names,
+        opset_version=17,
+        do_constant_folding=True,
+    )
+    print(f">>> ONNX exported to {save_path}")
+    xformers.ops.memory_efficient_attention = _orig_mea
+    return n_outputs
+
+
+class TRTBackbone:
+    """TensorRT runtime wrapper for the video UNet backbone."""
+
+    def __init__(self, engine_path, n_hs_a=9, device=None):
+        import tensorrt as trt
+
+        self.device = _normalize_cuda_device(device)
+        self.logger = trt.Logger(trt.Logger.WARNING)
+        with torch.cuda.device(self.device):
+            with open(engine_path, 'rb') as f:
+                runtime = trt.Runtime(self.logger)
+                self.engine = runtime.deserialize_cuda_engine(f.read())
+            self.context = self.engine.create_execution_context()
+        self.n_hs_a = n_hs_a
+
+        import numpy as np
+        self.output_buffers = {}
+        with torch.cuda.device(self.device):
+            for i in range(self.engine.num_io_tensors):
+                name = self.engine.get_tensor_name(i)
+                if self.engine.get_tensor_mode(name) == trt.TensorIOMode.OUTPUT:
+                    shape = self.engine.get_tensor_shape(name)
+                    np_dtype = trt.nptype(self.engine.get_tensor_dtype(name))
+                    buf = torch.empty(
+                        list(shape),
+                        dtype=torch.from_numpy(np.empty(0, dtype=np_dtype)).dtype,
+                        device=self.device)
+                    self.output_buffers[name] = buf
+                    print(
+                        f"    TRT output '{name}': {list(shape)} {buf.dtype} on {self.device}"
+                    )
+
+    def __call__(self, h, emb, context):
+        import numpy as np
+        import tensorrt as trt
+
+        bound_inputs = {}
+        with torch.cuda.device(self.device):
+            for name, tensor in [('h', h), ('emb', emb), ('context', context)]:
+                expected_dtype = trt.nptype(self.engine.get_tensor_dtype(name))
+                torch_expected = torch.from_numpy(
+                    np.empty(0, dtype=expected_dtype)).dtype
+                if tensor.device != self.device or tensor.dtype != torch_expected:
+                    tensor = tensor.to(device=self.device,
+                                       dtype=torch_expected,
+                                       non_blocking=True)
+                tensor = tensor.contiguous()
+                bound_inputs[name] = tensor
+                self.context.set_tensor_address(name, tensor.data_ptr())
+
+            for name, buf in self.output_buffers.items():
+                self.context.set_tensor_address(name, buf.data_ptr())
+
+            stream = torch.cuda.current_stream(device=self.device)
+            self.context.execute_async_v3(stream.cuda_stream)
+
+        y = self.output_buffers['y']
+        hs_a = [self.output_buffers[f'hs_a_{i}'] for i in range(self.n_hs_a)]
+        return y, hs_a

unitree_g1_pack_camera/case1/output/inference/analysis/analysis_report.md

@@ -0,0 +1,54 @@
+# DDIM Analysis Report
+
+## Analysis Config
+
+- Stability threshold: None
+- Stability window: 3
+- Mid-step window: 15-35
+
+## Dataset Overview
+
+- Unique samples: 1
+- Step rows: 1100
+- Sample rows: 2
+- Round rows: 11
+- Scenes: unitree_g1_pack_camera
+- Pass types: policy, world_model
+
+## Five Directions
+
+### Original Early Stop
+
+- `latent_first_stable_step_policy`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_first_stable_step_world_model`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_l2_vs_full50`: mean=172.5522, median=175.0850, p90=194.0716, p95=200.4957
+
+### Direction C: Action Converges First
+
+- `action_first_stable_step`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_first_stable_step`: mean=nan, median=nan, p90=nan, p95=nan
+- `action_vs_full50_95pct_step`: mean=47.2727, median=47.2727, p90=47.2727, p95=47.2727
+- `share_action_first_stable_before_latent`: nan
+- `share_action_95pct_before_latent`: nan
+
+### Direction D: Cross-Step Similarity
+
+- `latent_delta_mid_steps`: mean=0.3526, median=0.3486, p90=0.4168, p95=0.4280
+- `action_delta_mid_steps`: mean=0.0301, median=0.0287, p90=0.0431, p95=0.0455
+- `state_delta_mid_steps`: mean=0.0312, median=0.0299, p90=0.0444, p95=0.0465
+
+### Direction A: Budget Heterogeneity
+
+- `oracle_budget_action`: mean=46.8182, median=46.8182, p90=46.8182, p95=46.8182
+- `oracle_budget_state`: mean=46.5455, median=46.5455, p90=46.5455, p95=46.5455
+- `oracle_budget_action_by_scene`:
+  - `unitree_g1_pack_camera`: mean=46.8182, median=46.8182, std=0.0000
+
+### Direction B: Round Reuse
+
+- `latent_init_dist_to_prev_round`: mean=231.3224, median=234.0656, p90=246.2483, p95=251.1132
+- `action_drift_vs_prev_round`: mean=0.1699, median=0.1659, p90=0.2361, p95=0.3048
+- `round_total_time_s`: mean=136.4094, median=136.1244, p90=137.2721, p95=137.4062
+- `policy_pass_total_time_s`: mean=33.5494, median=33.5109, p90=33.7524, p95=33.8025
+- `world_model_pass_total_time_s`: mean=33.6186, median=33.5301, p90=33.9147, p95=34.0011
+

unitree_g1_pack_camera/case1/output/inference/analysis/scene_summary.csv

@@ -0,0 +1,4 @@
+scene,pass_type,pass_total_time_s,pass_total_time_s,pass_total_time_s,action_first_stable_step,action_first_stable_step,action_first_stable_step,state_first_stable_step,state_first_stable_step,state_first_stable_step,latent_first_stable_step,latent_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_90pct_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,action_vs_full50_95pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_action,oracle_budget_action,oracle_budget_state,oracle_budget_state,oracle_budget_state,round_total_time_s,round_total_time_s,round_total_time_s,policy_pass_total_time_s,policy_pass_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,world_model_pass_total_time_s,world_model_pass_total_time_s,latent_init_dist_to_prev_round,latent_init_dist_to_prev_round,latent_init_dist_to_prev_round,action_drift_vs_prev_round,action_drift_vs_prev_round,action_drift_vs_prev_round
+,,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std
+unitree_g1_pack_camera,policy,33.54937371340665,33.54937371340665,,,,,,,,,,,45.18181818181818,45.18181818181818,,47.27272727272727,47.27272727272727,,46.81818181818182,46.81818181818182,,46.63636363636363,46.63636363636363,,136.40941752086985,136.40941752086985,,33.54937371340665,33.54937371340665,,33.61860912496393,33.61860912496393,,231.32244110107425,231.32244110107425,,0.1698735475540161,0.1698735475540161,
+unitree_g1_pack_camera,world_model,33.61860912496393,33.61860912496393,,,,,,,,,,,45.6,45.6,,47.9,47.9,,46.81818181818182,46.81818181818182,,46.54545454545455,46.54545454545455,,136.40941752086985,136.40941752086985,,33.54937371340665,33.54937371340665,,33.61860912496393,33.61860912496393,,231.32244110107425,231.32244110107425,,0.1698735475540161,0.1698735475540161,

unitree_g1_pack_camera/case1/output/inference/analysis/step_aggregate.csv

@@ -0,0 +1,102 @@
+pass_type,step,step_time_s,step_time_s,step_time_s,latent_delta,latent_delta,latent_delta,action_delta,action_delta,action_delta,state_delta,state_delta,state_delta,action_cosine_vs_full50,action_cosine_vs_full50,action_cosine_vs_full50,state_cosine_vs_full50,state_cosine_vs_full50,state_cosine_vs_full50,latent_l2_vs_full50,latent_l2_vs_full50,latent_l2_vs_full50
+,,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std
+policy,1,0.6878367554057728,0.6880619525909424,0.0033110620523770797,1.414315397089178,1.4148871898651123,0.0020325428650545185,0.04114515435966572,0.0385311245918273,0.006729751254617576,0.03353392942385237,0.0324590355157852,0.005213655934941631,0.05104054273529482,0.0594044253230094,0.08061243367448104,0.04970239153639833,0.0510599464178085,0.0433968937741515,171.53487188165838,173.13885498046875,9.664412595735614
+policy,2,0.6476822332902388,0.6467409133911133,0.0033061949903357337,1.0219056389548562,1.0213992595672607,0.0021199227925337992,0.03117037869312541,0.0313579067587852,0.00128079844960254,0.03157321905547918,0.0311720389872789,0.001233561510136463,0.05620139464735983,0.0597818084061145,0.08012955603299711,0.04808159629729659,0.051007229834795,0.042785685991889186,170.7512831254439,172.5921630859375,8.884331795400193
+policy,3,0.6487778316844593,0.6475324630737305,0.003142656369118312,0.8511151563037526,0.8503890037536621,0.0022678079357062162,0.030918842012231953,0.0312048215419054,0.0013150623303982193,0.0314352546225894,0.0309654027223587,0.0012646352731930529,0.061267804015766474,0.0607415698468685,0.07969713662891846,0.046423285352912784,0.0509076341986656,0.042235624551005185,170.08949279785156,171.49826049804688,8.215684603325563
+policy,4,0.6487634832208807,0.6478748321533203,0.004029844748812948,0.7509499354795977,0.7515787482261658,0.002085653877243151,0.030670592392032747,0.0310523509979248,0.0013543947943310654,0.03129686889323317,0.0308115370571613,0.0012925313653133462,0.06623721884732894,0.0637962371110916,0.07931629750658403,0.04472967067902734,0.0506284832954406,0.0417506693755794,171.22090981223366,173.73974609375,9.969544995601543
+policy,5,0.6487428275021639,0.6482586860656738,0.003670227295705804,0.6813994104211981,0.6816902160644531,0.0012989316136776511,0.03042646480554879,0.0309010632336139,0.0013973579091322895,0.03115857053886756,0.030674846842885,0.0013169888755812213,0.07110742039301177,0.0669608265161514,0.07898798133724029,0.0430029126900163,0.0474095977842807,0.04133433106447461,171.7985049161044,173.38519287109375,10.522559323839568
+policy,6,0.6903667883439497,0.6897494792938232,0.004114674566720337,0.631840321150693,0.6317303776741028,0.0010384486047498446,0.017322231338105374,0.017490204423666,0.002910828001447578,0.015140305696563227,0.0140756284818053,0.0030282631265205373,0.07776769216765055,0.0715667605400085,0.07839767970195065,0.051973039961673954,0.0585936158895492,0.04022245959692242,172.16226612437856,174.35243225097656,10.048036575294242
+policy,7,0.647422963922674,0.6468949317932129,0.003583353413726638,0.5937905853444879,0.5938090682029724,0.001921592650431461,0.015154530971564972,0.014760804362595,0.0017363361167851978,0.013371904803947909,0.0134965674951672,0.0014619184424924496,0.08950046229768881,0.0832099169492721,0.07804459230640924,0.06229589049789033,0.0696425959467887,0.039320697540958315,171.9501523104581,175.45179748535156,11.009355847979394
+policy,8,0.6494349566372958,0.647907018661499,0.003678409129130053,0.5631717064163901,0.5630322098731995,0.0015817508860227778,0.015223807570609127,0.0147224944084882,0.0017880035632544026,0.013635912148112545,0.0137267243117094,0.0014837978406407751,0.10123876702379095,0.0948357582092285,0.07768505695281973,0.07267526875842699,0.0803469270467758,0.03844046737145004,173.17877890846947,175.42636108398438,11.4794563409528
+policy,9,0.6486549810929731,0.6476733684539795,0.003381958888224332,0.5368096720088612,0.5368063449859619,0.0012997505336453795,0.015351818281818473,0.0147450612857937,0.0018353458492489237,0.013964270326224182,0.0140183633193373,0.001500480533043123,0.11299719796939325,0.1064594984054565,0.07732050089950257,0.08312634243206542,0.0911136791110038,0.037584999423217455,173.57969665527344,173.2157745361328,11.529347511729522
+policy,10,0.648524674502286,0.6476876735687256,0.0027746995862040877,0.5145620866255327,0.5149137377738953,0.0009753303702358603,0.015537748536602964,0.0148282889276742,0.0018782466111277533,0.014354296028613999,0.0143318716436624,0.0015125187965698307,0.12479040107097136,0.1180964112281799,0.07695241277282897,0.09366465258327393,0.1019586473703384,0.036758026894356026,174.62435774369672,175.2415313720703,12.050185505145521
+policy,11,0.6898852478374135,0.6890826225280762,0.0030885671704322937,0.4960190111940557,0.4958719313144684,0.001043292073665813,0.015973689542575262,0.015349842607975,0.0020938418756725062,0.016390399076044518,0.0157551653683185,0.002706246595735368,0.1348416229540651,0.1277795135974884,0.07631651612990974,0.10531310025941236,0.1145065873861312,0.03581093023896857,175.1026375510476,176.33164978027344,13.177995177056488
+policy,12,0.6468431732871316,0.6457967758178711,0.003362720633531823,0.4789777262644334,0.4789286255836487,0.0009239884415968051,0.01604176659814332,0.0156049719080328,0.0019857656329405377,0.01613926032388747,0.0159319564700126,0.0017793809230182487,0.14789754101498556,0.140600100159645,0.07625366604667033,0.11782875521616498,0.1270077675580978,0.03520866802768808,175.61699745871803,176.5402374267578,13.653373380154273
+policy,13,0.6488430283286355,0.6479635238647461,0.0029849957723824872,0.46395906805992126,0.4641004204750061,0.0011387814047693577,0.0165945177220485,0.0160846132785081,0.0019796246714622327,0.01682665207507929,0.0166594311594963,0.0017492656512210626,0.16111562638120214,0.1535690426826477,0.07619493913097489,0.13055456836115226,0.1393173187971115,0.03470546772424351,175.6337294145064,177.35533142089844,14.564181269406705
+policy,14,0.6488324295390736,0.6481714248657227,0.0029079461146256755,0.4505833468653939,0.450753778219223,0.0006972833685256457,0.0171962983228943,0.0166154094040393,0.0019732276184840215,0.017560891129753772,0.0174339599907398,0.001718903092465139,0.1745160174640742,0.1667066812515258,0.0761408469891097,0.14351325008002191,0.151850163936615,0.03431108104226634,175.81662264737216,177.41856384277344,14.752261446057327
+policy,15,0.6492104747078635,0.6492857933044434,0.0030156032274872503,0.43780169974673877,0.4381426572799682,0.0014141717283896432,0.01784540272571819,0.0171954650431871,0.001967209596037606,0.018340792676264553,0.0182484574615955,0.0016887167037235795,0.18811950358477503,0.1800341308116912,0.07609191356905544,0.1567286374894055,0.1646297574043274,0.03403519989026083,176.52252474698153,178.44029235839844,15.270175006710678
+policy,16,0.6900773048400879,0.6890242099761963,0.002834005473950311,0.4265059937130321,0.426725685596466,0.0012757082261086554,0.019436938518827555,0.0184130184352397,0.0019102748162865856,0.019923606379465556,0.0194175690412521,0.0020080569400021484,0.20049659907817838,0.1925880163908004,0.07555931317292137,0.17054133388129145,0.1792239248752594,0.033599076763758104,176.93490878018466,178.7749481201172,15.25540329070601
+policy,17,0.6468632871454413,0.645941972732544,0.0029188517330911625,0.4155325510285117,0.4154058694839477,0.0009717310046405538,0.01983847469091409,0.0191336963325738,0.001807244924928824,0.02067851021208539,0.0200859084725379,0.0019198111127640444,0.2151951891454783,0.2070861905813217,0.07574010643267212,0.18524082140489054,0.1955353021621704,0.033596649278282945,177.29241249778053,179.13311767578125,15.684465050010646
+policy,18,0.6488500941883434,0.6483078002929688,0.002787265702773912,0.40558123317631806,0.4055310785770416,0.0012127242393607735,0.020679541270841203,0.0199570972472429,0.001789346126194033,0.021615736186504326,0.0210293941199779,0.0018806571585265745,0.2302071018652482,0.2215050458908081,0.0759234647173602,0.20031294091181318,0.2092854976654052,0.03380125338553311,177.76761280406606,179.76133728027344,16.13099088598484
+policy,19,0.6489606987346302,0.6480269432067871,0.002884621551526869,0.39586012200875714,0.3958815634250641,0.0014227815184279114,0.02156404355032873,0.0208283551037311,0.0017752316644639764,0.02259749834510408,0.0220185667276382,0.0018415614979522589,0.2455596422607248,0.2344343066215515,0.07610756649873072,0.21578920429403128,0.2233870923519134,0.03421933755583985,177.99620056152344,179.90280151367188,16.19181694056777
+policy,20,0.649939775466919,0.6487793922424316,0.003337778775041695,0.38741510293700476,0.3873323500156402,0.001070676477871489,0.022492310201579827,0.021743968129158,0.0017653017183111037,0.023624791679057164,0.0230544228106737,0.0018025117843486185,0.26128153367476026,0.247724175453186,0.07629027699796084,0.2317030402747067,0.2378720194101333,0.0348531048603225,178.57843711159447,180.65936279296875,16.554021378614078
+policy,21,0.6903411475094882,0.6891791820526123,0.0031084136440534937,0.3787730580026453,0.3790284991264343,0.00111180655914367,0.024358985098925465,0.0239272434264421,0.0016587383496431166,0.024875037710774973,0.024390373378992,0.001776400796567733,0.2759955051270398,0.2629846334457397,0.07588656458955229,0.24835584109479727,0.2533071041107178,0.03573891197668474,178.9801219593395,180.74676513671875,16.741849853133633
+policy,22,0.6478268666700884,0.647162675857544,0.0031794350496444678,0.3707492568276145,0.3708417415618896,0.0007747079227107497,0.024935555390336248,0.0243407282978296,0.0017259436820161336,0.026206061752005013,0.0256872083991765,0.001665194629441521,0.29293703761967743,0.280838131904602,0.0759656290426331,0.26578986915675074,0.2692347168922424,0.03693711113398246,179.20561356977984,181.45278930664065,16.726193147869648
+policy,23,0.6492330161007968,0.648576021194458,0.0024890905830428265,0.36270924589850684,0.3628093004226684,0.0009535796249857153,0.026030953973531664,0.0254298374056816,0.0017321721327890189,0.027389374307610716,0.0268772691488266,0.0016137853600784492,0.3103741095824675,0.2992132902145386,0.07603394396140588,0.28379647162827576,0.285689502954483,0.03831765065383119,179.41348682750356,182.05218505859372,16.888674539601187
+policy,24,0.6493923447348855,0.6488265991210938,0.0026554209746684656,0.35577589273452753,0.3557948470115661,0.0008387584919050787,0.027175211770967954,0.0266359224915504,0.001745462770351593,0.028625223616307382,0.028171295300126,0.0015619395604417581,0.3283431123603474,0.3181472420692444,0.0760863315740385,0.3024178323420611,0.3027148246765136,0.03986189664408115,179.70676214044744,182.22169494628903,16.939215220968265
+policy,25,0.6493374434384432,0.6487107276916504,0.002525203571209384,0.34857072884386237,0.3485554158687591,0.0008008479637073065,0.02837064625187349,0.0279321242123842,0.0017667057459145542,0.029916533861647908,0.029551038518548,0.001510145950669776,0.34688206152482465,0.3376784920692444,0.07611648558079218,0.32169816710732196,0.3214923441410064,0.041548365384970004,180.16854026100853,182.5570373535156,16.886035078009588
+policy,26,0.6902096488259055,0.6892485618591309,0.0029407657957045497,0.3411648571491241,0.3408058285713196,0.001123105345350985,0.030783612619746756,0.0304605718702077,0.001476674330146288,0.03140442124144592,0.0312412027269601,0.0014218353476336072,0.36479574170979584,0.3562508523464203,0.07573208810187618,0.34187686985189264,0.3407418727874756,0.04331933566129255,180.5151922052557,183.03707885742188,16.953149993606228
+policy,27,0.6476634415713224,0.6468138694763184,0.0024003060188102143,0.33519297567280854,0.3354537785053253,0.0010030481895208188,0.031751416623592314,0.0315451174974441,0.0017734033735767414,0.032884837720881764,0.0328904576599597,0.001370065939056996,0.38512093370611017,0.376745343208313,0.07567572570105585,0.36290589516813104,0.3588489294052124,0.045221554927604185,180.76152177290484,183.3007049560547,17.09625542708051
+policy,28,0.6495674740184437,0.6488912105560303,0.002430485980168429,0.32868456298654725,0.3287448585033417,0.0007878800637976771,0.03316272050142283,0.0331250093877315,0.0018339247636266058,0.034366418692198614,0.0346297249197959,0.001325378669462673,0.4061857218092138,0.3980055451393127,0.07556180334498909,0.38474978371099994,0.3789873123168945,0.047171386377754504,181.05065501819956,183.62962341308597,17.155510122349845
+policy,29,0.6494769399816339,0.6485552787780762,0.0029840179604763363,0.3223639916289936,0.3224748969078064,0.0007426131906017524,0.034637889062816396,0.034546323120594,0.0019106604098747624,0.03591593287207859,0.0361089818179607,0.0012928122842774013,0.4280338829213923,0.4200791716575622,0.07537381228398292,0.4074572351845828,0.4015395641326904,0.0491332034550296,181.26195040616122,183.69125366210935,17.10375037298991
+policy,30,0.6496965885162354,0.6489639282226562,0.002874314371060168,0.31617663123390893,0.3159787356853485,0.0008778682828864555,0.03617924858223304,0.0360241122543811,0.002006126763700036,0.03753605959090314,0.0376458279788494,0.001279899184348048,0.4507078392939134,0.4430129528045654,0.07509157547735983,0.4310755512931129,0.4276289343833923,0.05106696660139952,181.51826199618253,184.1066436767578,17.121577691801246
+policy,31,0.6906413815238259,0.689673900604248,0.0029908820306697886,0.31114687431942334,0.3108324110507965,0.0008910054263999438,0.03905006633563471,0.0389620512723922,0.0015285771968997941,0.03995414213700724,0.0396933406591415,0.0013200304412765956,0.4729691364548423,0.4662858545780182,0.0745689033336959,0.45614778995513916,0.4518061876296997,0.05277327566303436,178.2308710271662,180.93716430664065,17.768244307360863
+policy,32,0.6471490209752863,0.6464581489562988,0.0026989591537964932,0.30277256261218677,0.3026106357574463,0.0007491671859776976,0.04037484187971457,0.039942268282175,0.0023468613716485258,0.04178366166624153,0.0411935783922672,0.0013880781628493315,0.49808967926285486,0.4950447380542755,0.07401218948058365,0.4823830317367207,0.47747802734375,0.05448803471199863,175.4517128684304,178.54397583007812,18.472553414017202
+policy,33,0.64940391887318,0.6483368873596191,0.003502132192177385,0.2948594364252957,0.2950589656829834,0.0007756345344997465,0.042184686796231656,0.0417098365724086,0.002534621598018112,0.04367475584149356,0.0432122722268104,0.0015105112457011658,0.5241922411051664,0.522653341293335,0.07326043486824467,0.5096868899735537,0.5042842626571655,0.055999153214614565,172.81015153364703,176.1564178466797,19.126928014640967
+policy,34,0.6495789831334894,0.6486401557922363,0.004122430616953803,0.28676206686279987,0.2867019176483154,0.0006572887406274552,0.04407523674043737,0.0435781516134738,0.0027580400049730313,0.045650859109379986,0.0454197190701961,0.001697478822069019,0.5512965023517609,0.5478223562240601,0.07226840519619174,0.5380799147215757,0.5326485633850098,0.05723330610148372,170.34932361949575,173.95498657226562,19.900468696811288
+policy,35,0.6493036963722922,0.6478614807128906,0.0037986918078235267,0.2781880118630149,0.2783993482589721,0.0009927312721243868,0.04604764350435946,0.0455246232450008,0.003022253409185645,0.047713040966879186,0.0473272688686847,0.001955412593093011,0.5794068818742578,0.5750817060470581,0.07098321253617014,0.5675655251199548,0.5629444122314453,0.058107761641841775,168.0462313565341,171.9365692138672,20.684901677617
+policy,36,0.6901766386899081,0.6892445087432861,0.002730508433528608,0.26934711228717456,0.2690409719944,0.0008993556442434482,0.05031285976821721,0.0504337474703788,0.0028695562008213132,0.05069000477140596,0.0500893481075763,0.0020294582394430618,0.6074602929028597,0.6047841906547546,0.06936945593736998,0.5987466194412925,0.5953270792961121,0.05847871664625564,165.99228390780362,170.1469268798828,21.40478935355118
+policy,37,0.6475749015808105,0.6466829776763916,0.003104822133963663,0.2598760046742179,0.2598170042037964,0.0007940421504128103,0.05207324976270844,0.0514971688389778,0.003952222470407499,0.052906427871097166,0.0522405989468097,0.00253300295234527,0.6387888409874656,0.6370430588722229,0.06726089031102463,0.6311332095753063,0.6289960145950317,0.05844791883001238,164.23874386874112,168.4765625,22.156137745149586
+policy,38,0.6494037888266824,0.6481273174285889,0.0036441426450518864,0.2506225827065381,0.2508217096328735,0.0006884617595517046,0.05440256202762772,0.0539724081754684,0.004393491899245238,0.05524211410771711,0.0546240843832492,0.0030376307051590127,0.6710418950427662,0.6702436804771423,0.06463273927918842,0.664443010633642,0.6637422442436218,0.057716824167270066,162.72628714821556,166.9638214111328,22.862525433136362
+policy,39,0.6494896628639915,0.6489975452423096,0.0029879730012665248,0.24062424762682477,0.2404987961053848,0.0008774876995432513,0.05680657008832148,0.0565774329006671,0.004905476269521756,0.057647742331027936,0.0570824593305587,0.0036384425408808366,0.7040680809454485,0.7042167782783508,0.06140304792030519,0.6985130526802756,0.6993643045425415,0.05617446795778871,161.55195201526988,165.9738311767578,23.520533112233668
+policy,40,0.6497342803261497,0.6485631465911865,0.0029035859791836454,0.2308672585270621,0.2311120778322219,0.0008852260142695257,0.05927506495605812,0.059281725436449,0.0054952605020595615,0.06011211770502,0.0596033409237861,0.004339383219605359,0.7376474792307074,0.7387198209762573,0.05749683787818806,0.7331103845076128,0.7355687022209167,0.053722231406150706,160.59454345703125,165.13827514648438,24.190207706810018
+policy,41,0.6913177533583208,0.6902580261230469,0.002919322142143197,0.22082572362639682,0.2209605723619461,0.0006679720293644594,0.06522218408909707,0.0651715099811554,0.006074562481141115,0.0622590753165158,0.0622150115668773,0.004580057237697356,0.7714574553749778,0.7740613222122192,0.052925006736720924,0.7676653265953064,0.7718186378479004,0.05024467643451416,159.87384726784447,164.81886291503906,24.861275236465836
+policy,42,0.6480572440407493,0.6468923091888428,0.002816798339386263,0.21070447157729752,0.2108505070209503,0.0010642295608918881,0.06769390709020871,0.066586323082447,0.007353183625993495,0.06443566964431237,0.0640224739909172,0.0057775792360396806,0.8069780198010531,0.8104296922683716,0.047207557146698914,0.8021321946924383,0.8066082000732422,0.045904697626584355,159.46489507501775,164.8072052001953,25.484337577151102
+policy,43,0.6498769196597013,0.6493535041809082,0.00292764946725251,0.20067393237894227,0.2008374780416488,0.0008326780693983284,0.07044370336966077,0.0691365003585815,0.008279990738379909,0.06685433299704027,0.0664953142404556,0.006758116806768652,0.841712848706679,0.8458664417266846,0.04073048438497873,0.8358337770808827,0.8400559425354004,0.04061626926532008,159.4006146517667,164.99317932128906,26.05717363064473
+policy,44,0.6497088345614347,0.6483900547027588,0.0031458799493430446,0.1901078738949515,0.1900599747896194,0.000835779413660312,0.0731296576559543,0.0715932622551918,0.0092876631544886,0.0692144811830737,0.0689045712351799,0.007821781709236341,0.8748954263600436,0.879603385925293,0.033676642521734725,0.8681361187588085,0.8722075819969177,0.03453149163221853,159.6634854403409,165.3663330078125,26.52328281544532
+policy,45,0.6493879448283802,0.6479225158691406,0.00308306739875395,0.17993601614778687,0.1798987239599228,0.0006745431973834978,0.07569428262385453,0.0739035010337829,0.010352734868978685,0.07147307727824553,0.0712060555815696,0.00894830349404384,0.9056182557886298,0.9112069606781006,0.026393829207304036,0.8982963236895475,0.902340054512024,0.02793602556798865,160.19171350652522,166.07191467285156,27.000039302783943
+policy,46,0.6916682069951837,0.6901402473449707,0.003235831745303403,0.1694915795868093,0.1695273071527481,0.0007676179605943142,0.0749267380345951,0.0751013755798339,0.010186926620350648,0.07493788850578391,0.0746354609727859,0.009433286246587015,0.9314036531881853,0.939341366291046,0.019981248448500237,0.9260419282046232,0.9307255148887634,0.02113542911136927,161.05496285178444,167.27197265625,27.399960883240936
+policy,47,0.6479458158666437,0.6470165252685547,0.002939416292823484,0.1590046801350333,0.1589466780424118,0.0007104769573467058,0.07608906856992025,0.0749063640832901,0.011430892817241807,0.07627307894555002,0.0762217193841934,0.010903964098569393,0.9535475427454169,0.9582948684692384,0.0144273704887614,0.9497837261720137,0.9545594453811646,0.01497443408472073,162.2230522849343,168.61753845214844,27.705098786754046
+policy,48,0.6498834003101696,0.6487264633178711,0.003154755192489887,0.1481544118035923,0.1481597125530243,0.0008805048277245049,0.07783281261270691,0.0766805186867713,0.01233740032171389,0.07793417979370461,0.0781041607260704,0.011973086220718888,0.970752705227245,0.973758101463318,0.010781227572093467,0.9686193628744646,0.9717366099357604,0.010261608100136316,163.75716400146484,170.35040283203125,27.931588621865217
+policy,49,0.6498605121265758,0.649031400680542,0.002414723186341074,0.13615625961260358,0.136216327548027,0.000985854537190554,0.07997240634127092,0.0788784250617027,0.013240963689978777,0.07998684251850298,0.0803201794624328,0.013035539360231862,0.9823288538239219,0.9862905740737916,0.009593597736915604,0.9818726236169989,0.9832581877708436,0.007810424126442877,165.54673489657316,172.11949157714844,28.108213016765507
+policy,50,0.6502033363689076,0.6492252349853516,0.002668080759458149,0.11278119683265682,0.1126245632767677,0.001530410910874375,0.10536787862127474,0.1041062250733375,0.018020167120615497,0.10526390847834671,0.1057901456952095,0.017970800312776296,0.9877927790988575,0.9914222955703736,0.010104626003882629,0.98955622586337,0.9917941093444824,0.007151228649300022,166.18099975585938,172.7540283203125,28.036243397578065
+world_model,1,0.6901874325492166,0.6882033348083496,0.004652919347280704,1.4144444032148882,1.4147138595581057,0.001778954603618504,0.043042470447041736,0.0425725057721138,0.003978204993228438,0.03167057359083126,0.0306400023400783,0.006528327051636371,0.03900951633907177,0.001696488354355,0.07760349928077023,0.059067137708718104,0.0651557892560958,0.06335430425863327,173.1571377840909,174.72938537597656,10.798609326400124
+world_model,2,0.64943697235801,0.6468534469604492,0.004230046387044571,1.0218601443550803,1.021780490875244,0.0022755514142162913,0.030987820842049298,0.0309080425649881,0.0011576468557423889,0.03165003200146281,0.0317809171974659,0.0010746483491121,0.044054598801515306,0.0089030843228101,0.07676945460567107,0.05587455926632336,0.0612085647881031,0.06400958327006191,173.2716633189808,174.24234008789062,10.726102208896405
+world_model,3,0.6504674391313032,0.6489086151123047,0.0031923487264945566,0.8506394895640287,0.8501879572868347,0.002498805397997276,0.030692000450058373,0.0306419841945171,0.001150930191315527,0.03157710419459772,0.0316162593662738,0.001067722114161465,0.049015924673188764,0.0180806014686822,0.07601087567119799,0.05263608076017008,0.0573555119335651,0.06471389963057243,173.1281502463601,173.8313751220703,10.675043963609205
+world_model,4,0.6504956808957186,0.6495752334594727,0.004435427839873687,0.7497467994689941,0.7494067549705505,0.001885099283272064,0.030400894074277398,0.0303801465779542,0.0011466136782498308,0.031503169035369656,0.0314200334250927,0.0010719053517325428,0.05389132672412826,0.0230652745813131,0.07533151405872308,0.04935506942935963,0.055531982332468,0.06546415859501178,173.3203457919034,174.68341064453125,10.530761371728792
+world_model,5,0.6519384167411111,0.6487352848052979,0.008784575338937924,0.6828935579820112,0.6830169558525085,0.0011324579624125263,0.030115480111403845,0.0301234535872936,0.0011445476281554763,0.03142848932607602,0.031170155853033,0.001086612430548704,0.05867894027720794,0.0300141181796789,0.07473461212207674,0.046034867451949534,0.0536619536578655,0.0662570738064663,172.7358925559304,175.10391235351562,10.195742330893218
+world_model,6,0.6916962536898527,0.6903371810913086,0.004154800198986271,0.6321003166112033,0.6320990324020386,0.0014948292169425904,0.01992043023082338,0.0190781038254499,0.003440813872932783,0.014101324179633055,0.0135584389790892,0.002825846279772188,0.06412349353459747,0.0351062268018722,0.07408031715724238,0.055416586851193106,0.0637934654951095,0.06644833881915045,173.15597672895953,173.88002014160156,11.49105588111728
+world_model,7,0.6514237794009122,0.6475827693939209,0.00915957421144447,0.5945998592810198,0.5942923426628113,0.0011197081030958906,0.01567267008464441,0.0157627090811729,0.0013339951179554648,0.013397558507594137,0.0136717669665813,0.0012290416355712117,0.07656896977939386,0.0495465248823165,0.07326557490855556,0.0656305488876321,0.0742959082126617,0.0670829516639668,173.54277870871803,175.62802124023438,12.102565388760077
+world_model,8,0.651599797335538,0.6502435207366943,0.004465187652103753,0.5632444240830161,0.5637466907501221,0.001642459235826042,0.015729984268546052,0.0158609077334404,0.0013579440543655906,0.013672495514831691,0.0139606418088078,0.0012550559642339326,0.08901839703321456,0.0640158653259277,0.07245098206192592,0.07590023512867362,0.0848560407757759,0.06771930026833282,174.06075911088422,177.594970703125,12.81668476128768
+world_model,9,0.6503221121701327,0.6483035087585449,0.00394611930239549,0.5374385877089067,0.537429690361023,0.0010882228710769957,0.015843427655371726,0.0160141997039318,0.00137834746395476,0.014011308804831653,0.0143113322556018,0.001277091414991707,0.10148660669272593,0.078529343008995,0.07163759600513812,0.08624074303291059,0.0954893827438354,0.068357490996363,174.7043318314986,177.4236297607422,13.240369885566592
+world_model,10,0.6511294408278032,0.6494760513305664,0.004466187074867514,0.5158192948861555,0.5159617066383362,0.0007156410888342359,0.016012337228113918,0.016221722587943,0.0013951341112484726,0.014411329291760873,0.0147213311865925,0.0012955235631831995,0.11398845673962071,0.0931021049618721,0.07082644208648478,0.09666753009977663,0.1062117516994476,0.06899758086919397,174.98081831498578,176.79722595214844,13.748503238600245
+world_model,11,0.6926611770283092,0.6910960674285889,0.004765978778293706,0.49647371064532886,0.4969973266124725,0.0012595421708653074,0.016014237718825944,0.0155232986435294,0.0019479470361827096,0.015218021856112852,0.0152850402519106,0.0012158668903204902,0.12328078767115415,0.1033498272299766,0.06973828486809096,0.10852076519619333,0.1195420771837234,0.06983189622952582,175.4960049715909,176.7800750732422,14.098786651353414
+world_model,12,0.6490212353793058,0.6473500728607178,0.0046160324186714915,0.47975305264646356,0.479565680027008,0.000983023417399217,0.016255085017870728,0.0159771032631397,0.0018039424154487546,0.016049389812079283,0.0162255633622407,0.0012769541323290284,0.13653382252563126,0.1186997592449188,0.06886128721553969,0.12079445509748023,0.133454144001007,0.07082844102221672,175.78199490633878,177.07711791992188,13.995580324366617
+world_model,13,0.6508976546200839,0.6495089530944824,0.003978984905860109,0.4651044282046231,0.464564710855484,0.0008731037793563493,0.0168016333640976,0.016532514244318,0.0017975140619133749,0.016737388954921155,0.0169070549309253,0.0012535945030258161,0.14995464513247658,0.1333168298006057,0.06798987174642984,0.1332714992829344,0.1476023495197296,0.07183676598933214,176.30487337979403,177.14598083496094,14.161034502068382
+world_model,14,0.651021350513805,0.6492784023284912,0.00417389170577819,0.45156435803933576,0.451289027929306,0.0010239172803404386,0.017396829599006566,0.0171370506286621,0.0017882292591205732,0.01747254603965712,0.0176357757300138,0.0012317811627473353,0.16356341269883237,0.1478048861026764,0.06712589519013175,0.14597439088604663,0.1620111018419265,0.07285679687182879,177.33421603116122,178.4777069091797,14.506910432471836
+world_model,15,0.6514927040446888,0.6495330333709717,0.004664919674819356,0.43914141980084503,0.4388825595378876,0.0012084366692972732,0.018039025535637618,0.0177891328930854,0.001776654055652624,0.018253688107837264,0.018389493227005,0.0012118706739798419,0.17738096348263996,0.16253262758255,0.0662714072160867,0.15892669253728603,0.1767057478427887,0.07388824690876285,177.9435605135831,178.3133087158203,14.389846144778158
+world_model,16,0.6921203786676581,0.6901638507843018,0.004425738830520711,0.42715042558583344,0.4271185100078583,0.0008272944048967746,0.018957146697423616,0.0189746841788291,0.001498648788836524,0.019479272209785157,0.0194770786911249,0.0009772634953120307,0.18914962492205875,0.1755973994731903,0.06496890421837022,0.17269177192991425,0.1921688914299011,0.07495577865671992,178.27986561168325,179.6419219970703,14.462559825572567
+world_model,17,0.6484122493050315,0.6460320949554443,0.004296386707418081,0.41594838283278723,0.415938138961792,0.0009214414139512012,0.019813147627494518,0.0196120236068964,0.0017743326400395836,0.020460352809591683,0.0205086171627044,0.0010469377252662058,0.20365317843177097,0.1912849843502044,0.06413681996132752,0.1869080635634335,0.2080999016761779,0.07607767000480897,178.73463162508878,179.58245849609375,14.739052497443453
+world_model,18,0.6508032625371759,0.6495945453643799,0.003819534815310205,0.4063236713409424,0.4062586426734924,0.000931723349932183,0.020658364180814084,0.0204683411866426,0.001762333306343927,0.02140322073616761,0.0215877126902341,0.0010213025852475292,0.21847447482022367,0.207335963845253,0.0633144892118595,0.20148545842279084,0.224435344338417,0.0772040444334544,179.13285411487925,180.6522521972656,14.534487619909434
+world_model,19,0.6513201756910845,0.6489527225494385,0.004753040583758849,0.396343317898837,0.3966798186302185,0.0011427382891683092,0.02154712345112449,0.0213688928633928,0.0017502850231979172,0.02239140546457329,0.0225701667368412,0.000999235724318648,0.23364094306122168,0.2237806022167205,0.06250476295066697,0.21645567024295975,0.2412082999944687,0.0783326602809604,179.4692285711115,181.57687377929688,14.830400773900044
+world_model,20,0.6512038274244829,0.6492171287536621,0.0035241759244167567,0.3871028423309326,0.3873156309127807,0.00098534694240042,0.0224797219376672,0.0223140642046928,0.001738350435942084,0.023425992916930764,0.0235063470900058,0.0009812331560992568,0.24918148598887704,0.2406506836414337,0.061710719900407586,0.23185241019183936,0.2564210593700409,0.07946065303185088,179.9445662064986,182.00396728515625,14.985627613839792
+world_model,21,0.6920438246293501,0.6904504299163818,0.0033521634786565055,0.37909806858409534,0.3790543675422668,0.000662005945992562,0.023973630233244416,0.0243335850536823,0.0015613968128082828,0.02492396973750803,0.0246109440922737,0.0008161491990360672,0.2632763724435459,0.2563226819038391,0.06039404896579807,0.24817213483832096,0.2716187238693237,0.08055065875160423,180.17008556019175,182.74366760253903,15.208338840231294
+world_model,22,0.6486902236938477,0.6467409133911133,0.004426416272389367,0.37102228132161225,0.3710577487945556,0.0007112047501616492,0.02492232857780018,0.0250040404498577,0.0016613464356976786,0.0260381415824998,0.0257781613618135,0.0009125457836523909,0.27988873422145844,0.276051789522171,0.05964826942260305,0.2650917592373761,0.2878725528717041,0.08164315434321989,180.48986261541194,183.2282257080078,15.28474241139826
+world_model,23,0.6506305391138251,0.6497409343719482,0.003678865565155595,0.3630618507211858,0.3629414737224579,0.000739379388285571,0.026030991565097413,0.0261344946920871,0.0016535228955166035,0.027229785241863872,0.0268917195498943,0.0009218540142657063,0.2970033897594972,0.2970390021800995,0.05892153254695829,0.28257513723590155,0.3046449422836303,0.08270761835803979,180.9643318869851,183.6927947998047,15.327454201842684
+world_model,24,0.6501042192632501,0.6485157012939453,0.004412242207109715,0.3559333438223058,0.3560895919799804,0.0006849151175425958,0.02718895622952413,0.0273158084601163,0.0016464376430176764,0.028475327247923024,0.0283295083791017,0.0009418130117502714,0.314657600088553,0.3186452686786651,0.05821690741144442,0.3006658526984128,0.3219772577285766,0.0837349862868628,181.34929171475497,183.8955535888672,15.549992762167456
+world_model,25,0.6498888405886564,0.6486413478851318,0.004188039983681262,0.3484781357375058,0.3486402928829193,0.0009204351598956987,0.028398586775768828,0.0285505354404449,0.001640198057710302,0.02977787178348407,0.0297850407660007,0.0009740832549414689,0.33289024504748255,0.3409040868282318,0.057537203107036486,0.31940985267812555,0.3399132490158081,0.08471417995275903,181.6707333651456,183.98110961914065,15.802671204779184
+world_model,26,0.6914587237618186,0.6902298927307129,0.0036559824202188314,0.3418886200948195,0.3420698046684265,0.0009689214662833059,0.030680478804490728,0.0311465431004762,0.0016409113306172982,0.031634729524904974,0.0319922268390655,0.0009226283306849445,0.35001387379386206,0.3578331172466278,0.0563785932852537,0.33911974999037653,0.357910007238388,0.08563804260657316,181.98832980069247,184.13699340820312,16.022321182730078
+world_model,27,0.6478783434087579,0.6467366218566895,0.0036073823505274887,0.3350510895252228,0.3354585468769073,0.0009173198970796328,0.031836357814344456,0.032314296811819,0.0016523222699629785,0.032887480814348516,0.0328883305191993,0.0010673844369938612,0.36998876387422736,0.3775975704193115,0.055756329213058044,0.3597058450633829,0.3777479231357574,0.08650356651191735,182.23470653187144,184.47671508789065,16.047025396208905
+world_model,28,0.6509945609352805,0.6500871181488037,0.003932073620502374,0.3285938934846358,0.3285699486732483,0.0007882819735415304,0.033268089321526584,0.0337352082133293,0.0016524394955392504,0.034386295147917446,0.0343540087342262,0.0011636412894412797,0.390716092153029,0.398134171962738,0.05517291691502459,0.38111103664744983,0.398355484008789,0.08725937282294621,182.47592163085938,184.53407287597656,16.052746792997453
+world_model,29,0.6502645232460715,0.6494972705841064,0.003783561745758387,0.32244389165531506,0.3224455416202545,0.0007439401229169474,0.03476468033411282,0.035096738487482,0.0016558561080097004,0.035955367440527075,0.0358854196965694,0.0012830078149553413,0.4122411771254106,0.4194912016391754,0.05462749694046213,0.4033876074986024,0.4197828769683838,0.08787906566297667,182.63534545898438,184.54876708984372,16.07287436054623
+world_model,30,0.6503014781258323,0.6494412422180176,0.003970907222283625,0.31588365002111957,0.315897136926651,0.0006702043632716315,0.0363285047086802,0.0369062460958957,0.0016639767392271309,0.03759760443459854,0.0374845415353775,0.0014276669800880747,0.43460802327502857,0.4419646859169006,0.05411630115797322,0.42658709666945716,0.4420797228813171,0.08833103154988454,182.98888882723722,184.89682006835935,16.108693990951604
+world_model,31,0.6913770545612682,0.6898956298828125,0.0029226032160639707,0.31085004047913983,0.3105895221233368,0.0006886395424621019,0.03904122825373297,0.0391105525195598,0.001891096978158309,0.040325681255622296,0.040342129766941,0.0016896293417133614,0.45602939887480304,0.4658563733100891,0.0532423211726477,0.45128933408043603,0.4651739001274109,0.08848367408324853,179.76125127618963,181.8824920654297,16.738872242675924
+world_model,32,0.6483270471746271,0.6466741561889648,0.004139776852614292,0.3031195510517467,0.3033312559127807,0.0008587066987933051,0.040560835125771404,0.0412463806569576,0.0016510197147745732,0.04194737903096456,0.0419787466526031,0.001940035755833772,0.4808616719462655,0.4926540851593017,0.0528206598437397,0.47717574509707367,0.4904298186302185,0.0883809867403788,176.90028103915128,179.12322998046875,17.514466104944635
+world_model,33,0.6512812050906095,0.6498806476593018,0.003775569166801228,0.2952803617173975,0.2951521277427673,0.0007807643409953634,0.042398944158445674,0.0430708341300487,0.001697799291566186,0.04387211054563518,0.043762557208538,0.0021998620622522783,0.5067003504796461,0.5183749198913574,0.052390613140716136,0.5041624226353385,0.5167524814605713,0.08794891632692776,174.28900423916903,176.6444854736328,18.2327954745758
+world_model,34,0.6500619541515004,0.6496114730834961,0.003149429592741486,0.2867125110192732,0.2866807281970978,0.000871037855587926,0.04431910880587313,0.0449702888727188,0.0017671329498104176,0.04588544571941545,0.0457460470497608,0.0025008229106638505,0.5335667973214929,0.5433066487312317,0.05192486219158705,0.5322764299132607,0.5441693663597107,0.08712248358387474,171.85226717862216,174.2517852783203,18.966612634847934
+world_model,35,0.651204065843062,0.6496100425720215,0.00445763119790942,0.27814969420433044,0.2783681750297546,0.0011477038191907646,0.04632234742695631,0.0469437316060066,0.001866606960002601,0.04798847335306076,0.0476731285452842,0.002848013499742028,0.5614666044712067,0.5693179965019226,0.05138756173529562,0.5615276992321014,0.5726915597915649,0.08582810872924633,169.67762062766334,172.3053436279297,19.63125334632274
+world_model,36,0.6926375519145619,0.6910841464996338,0.004842805677163122,0.2690770002928647,0.2689660787582397,0.0011502644756443885,0.05051623420281839,0.0500565916299819,0.002456158200035607,0.05126428942788727,0.0507188141345977,0.003393897903287729,0.5884867635640231,0.593462347984314,0.050464612519380274,0.5925961814143441,0.6025053858757019,0.08384996336873342,167.6769131747159,170.16061401367188,20.334496041932443
+world_model,37,0.6478479992259633,0.6466608047485352,0.004674231251787187,0.25988766280087555,0.2598007023334503,0.0009526101384487606,0.05250764434987845,0.0522770099341869,0.002336364287723352,0.05337200618603008,0.0526606328785419,0.003912944942415891,0.61965796622363,0.6227871775627136,0.04958375515522425,0.6248678727583452,0.6343057155609131,0.08119449739670587,165.98271595348012,168.48550415039062,20.98266467301096
+world_model,38,0.6510405757210471,0.6497147083282471,0.004311771208760232,0.2503832917321812,0.2505967915058136,0.0010678877965522413,0.054886226965622426,0.0543760135769844,0.0026207230000443306,0.05576236105777996,0.0546589940786361,0.004473018080102227,0.6517979069189592,0.6532410979270935,0.04844635483470794,0.6581191989508542,0.6691584587097168,0.07776793144556893,164.50430297851562,167.13604736328125,21.55986656942007
+world_model,39,0.6504490375518799,0.6488654613494873,0.00465336713642716,0.2410306171937422,0.2410803884267807,0.0012195293454732511,0.05734118175777518,0.0570101998746395,0.0029798605506222544,0.058223324743184125,0.0566954649984836,0.005106405490399932,0.6847539002245123,0.6847182512283325,0.04698803031979391,0.6921813813122836,0.7046986818313599,0.0734948564404534,163.3295863758434,166.0323944091797,22.17769809387432
+world_model,40,0.6508629322052002,0.6494452953338623,0.004576488606697954,0.23136020790446885,0.2314299494028091,0.0011789156132302387,0.05986166034232481,0.0597445629537105,0.003423388988615888,0.06074120138179168,0.059157159179449,0.005816641176360152,0.7183032090013678,0.7170422673225403,0.04515432654704915,0.7268075021830472,0.7396231889724731,0.06832905798583494,162.49524619362572,165.2713165283203,22.781019752970263
+world_model,41,0.6922155727039684,0.6907162666320801,0.005327708608661885,0.22104071487079963,0.2210594266653061,0.0010287715528434784,0.06540149416435845,0.0652457475662231,0.004108399542781783,0.06320663643154226,0.0617449954152107,0.006528506464661646,0.7518102797594938,0.7472197413444519,0.042713778221503663,0.7615116401152178,0.7723662853240967,0.06238410166086222,161.82319779829547,164.75518798828125,23.346397796000105
+world_model,42,0.6481400403109464,0.6467063426971436,0.00416305320520217,0.21083427830175916,0.2111102193593979,0.0009207252469237854,0.06847119128162206,0.0685188323259353,0.004980954938150018,0.06526798924261869,0.0640973523259162,0.007421993195521065,0.7874898747964338,0.7833424210548401,0.039931594138619984,0.7960487929257479,0.8045583963394165,0.05564672385760426,161.4727644486861,164.37705993652344,23.806287017553668
+world_model,43,0.6509171182459051,0.6495156288146973,0.004354069611373433,0.20052815296433185,0.2004192471504211,0.0010224929966304527,0.07127646560018709,0.0707914903759956,0.0058097854239850715,0.06770771470936858,0.0663956403732299,0.008346520449957819,0.8224140080538663,0.8203790187835693,0.036920341846987344,0.8297322338277643,0.8364346623420715,0.048373623769245803,161.44782811945134,164.54066467285156,24.256224997489362
+world_model,44,0.6500566005706787,0.6484003067016602,0.004086679311997546,0.19050529734654856,0.1902331262826919,0.000901442496473915,0.07401509989391669,0.0728944316506385,0.006744511523976642,0.07006411762400101,0.0686012133955955,0.009308563220881968,0.8558103984052484,0.8553849458694458,0.03401178281120753,0.8618640032681552,0.8685265183448792,0.041010247908861194,161.74714105779475,165.0198974609375,24.601765063744907
+world_model,45,0.6503547321666371,0.6487326622009277,0.004700370480834122,0.17996519397605545,0.1799974590539932,0.0009289944389267497,0.07662914354692801,0.0749488845467567,0.007768473512344501,0.07228585061701859,0.0706079006195068,0.010275269597115359,0.8867689642039213,0.888229489326477,0.03170456924367929,0.8916336948221381,0.8983480334281921,0.03426875212464518,162.36859893798828,165.87425231933594,24.893218370361684
+world_model,46,0.691618182442405,0.6904275417327881,0.004725710850974771,0.1696115136146545,0.1696838736534118,0.0010091499533785404,0.07581636174158612,0.0732993558049202,0.008170244813517805,0.07570562850345258,0.0727193877100944,0.011004059569505243,0.9128518538041548,0.9195061326026917,0.030976186907571784,0.918533287265084,0.9262393712997437,0.029459714182489383,163.2966905073686,166.87924194335938,25.194325275221352
+world_model,47,0.6479579968885942,0.6464476585388184,0.004323917106301669,0.15929173339496955,0.159085527062416,0.0008048242902906626,0.07712150771509513,0.0743104144930839,0.009247555008462067,0.07666987756436518,0.0743667036294937,0.012088780229163869,0.9352076378735629,0.9455254077911376,0.031343893825378395,0.9410948265682567,0.9496381878852844,0.027046321106446596,164.57086181640625,168.3689727783203,25.4670827584787
+world_model,48,0.651075623252175,0.6494584083557129,0.004167791087546613,0.1482306840744885,0.1482225209474563,0.0009056038922777691,0.07889709215272551,0.0757696852087974,0.0102292972796746,0.07816946912895543,0.0757470726966857,0.012819184107163175,0.9526306770064614,0.958879828453064,0.03266242221851617,0.9585093801671808,0.9647461771965028,0.026975371693270522,166.07403841885653,170.1331024169922,25.671047711963098
+world_model,49,0.6511041467840021,0.649712085723877,0.003939582808815997,0.13609897142106833,0.1358009129762649,0.0010175619678860735,0.08107459748333147,0.0776293575763702,0.011221562530051351,0.08002151210199697,0.0775212123990058,0.013456395896265108,0.9644263982772827,0.97551429271698,0.03425955986493581,0.9701669432900168,0.9802457094192504,0.02807883866100148,167.90835571289062,172.18356323242188,25.835531533885295
+world_model,50,0.6502515619451349,0.6488831043243408,0.004427051714051182,0.11297817121852524,0.112374871969223,0.0014212208795290943,0.10683172128417272,0.1020421907305717,0.015577440477436321,0.1050022108988328,0.1017838791012764,0.01783347360950816,0.9701540686867454,0.9815363883972168,0.03557818470596488,0.9757647351785139,0.985155463218689,0.029247518631041223,168.50939802689985,172.75628662109375,25.788767103637852

unitree_g1_pack_camera/case1/output/inference/round_summary.csv

@@ -0,0 +1,12 @@
+sample_id,scene,round_id,policy_pass_total_time_s,world_model_pass_total_time_s,round_total_time_s,latent_init_dist_to_prev_round,action_drift_vs_prev_round,psnr_full50
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,0,33.611624240875244,34.08755350112915,137.2721483707428,,,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,1,33.85250473022461,33.91472029685974,137.54015183448792,214.96084594726562,0.18166154623031616,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,2,33.75244617462158,33.62014126777649,137.18953275680542,240.2042236328125,0.19001662731170654,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,3,33.52527594566345,33.530078172683716,136.1212797164917,206.36373901367188,0.02651798725128174,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,4,33.39732813835144,33.489410638809204,135.91671419143677,216.5181121826172,0.1665964126586914,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,5,33.47249174118042,33.59989905357361,136.12437462806702,228.9165802001953,0.373574435710907,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,6,33.53043866157532,33.48976278305054,136.1382839679718,239.2146759033203,0.2208324670791626,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,7,33.510878801345825,33.56129765510559,136.194073677063,245.16720581054688,0.13505703210830688,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,8,33.4688138961792,33.51669669151306,136.05504417419434,255.97816467285156,0.1348034143447876,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,9,33.46222972869873,33.503395318984985,136.02690505981445,243.15744018554688,0.10453599691390991,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,10,33.459078788757324,33.49174499511719,135.9250843524933,222.74342346191406,0.1651395559310913,

unitree_g1_pack_camera/case1/output/inference/sample_summary.csv

@@ -0,0 +1,3 @@
+sample_id,scene,pass_type,pass_total_time_s,action_first_stable_step,state_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_state,oracle_budget_latent,latent_init_dist_to_prev_round,action_drift_vs_prev_round,round_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,psnr_full50
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,33.54937371340665,,,,45.18181818181818,47.27272727272727,46.81818181818182,46.63636363636363,,231.32244110107422,0.16987354755401612,136.40941752086985,33.54937371340665,33.61860912496393,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,33.61860912496393,,,,45.6,47.9,46.81818181818182,46.54545454545455,,231.32244110107422,0.16987354755401612,136.40941752086985,33.54937371340665,33.61860912496393,

unitree_g1_pack_camera/case1/output/sparse_8/inference/round_summary.csv

@@ -0,0 +1,12 @@
+sample_id,scene,round_id,policy_pass_total_time_s,world_model_pass_total_time_s,round_total_time_s,latent_init_dist_to_prev_round,action_drift_vs_prev_round,psnr_full50
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,0,35.013712882995605,34.73551678657532,143.6849582195282,,,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,1,35.05562353134155,34.18100881576538,143.12280178070068,214.95654296875,0.1811339259147644,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,2,34.80648851394653,34.815757036209106,143.93157863616943,240.22900390625,0.18993812799453735,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,3,35.42078423500061,34.078826665878296,143.7456750869751,206.12351989746094,0.026802361011505127,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,4,34.88056492805481,34.83683705329895,145.21864223480225,216.09486389160156,0.16485852003097534,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,5,34.707502365112305,34.06549119949341,142.62004137039185,227.89527893066406,0.3844761848449707,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,6,34.08131694793701,34.453927993774414,141.4016900062561,236.65484619140625,0.21243447065353394,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,7,35.12070393562317,35.34656286239624,145.56022691726685,246.84864807128906,0.1723371148109436,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,8,34.99006509780884,34.1045401096344,142.48975276947021,236.0821533203125,0.10035890340805054,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,9,34.99584174156189,35.086113691329956,143.1674451828003,235.0965576171875,0.1056985855102539,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,10,35.301815032958984,34.37025237083435,144.55097150802612,241.2891845703125,0.16251075267791748,

unitree_g1_pack_camera/case1/output/sparse_8/inference/sample_summary.csv

@@ -0,0 +1,3 @@
+sample_id,scene,pass_type,pass_total_time_s,action_first_stable_step,state_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_state,oracle_budget_latent,latent_init_dist_to_prev_round,action_drift_vs_prev_round,round_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,psnr_full50
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,policy,34.943129019303754,,,,45.36363636363637,47.36363636363637,46.81818181818182,46.72727272727273,,230.12705993652344,0.17005489468574525,143.59034397385338,34.943129019303754,34.55225768956271,
+unitree_g1_pack_camera-vid0-fs6,unitree_g1_pack_camera,world_model,34.55225768956271,,,,45.9,48.0,46.81818181818182,46.54545454545455,,230.12705993652344,0.17005489468574525,143.59034397385338,34.943129019303754,34.55225768956271,

unitree_g1_pack_camera/case1/run_world_model_interaction.sh

@@ -0,0 +1,29 @@
+res_dir="unitree_g1_pack_camera/case1"
+dataset="unitree_g1_pack_camera"
+
+{
+    time CUDA_VISIBLE_DEVICES=0 "${PYTHON_BIN:-python}" scripts/evaluation/world_model_interaction.py \
+        --seed 123 \
+        --ckpt_path ckpts/unifolm_wma_dual.ckpt \
+        --config configs/inference/world_model_interaction.yaml \
+        --savedir "${res_dir}/output/sparse_8" \
+        --bs 1 --height 320 --width 512 \
+        --unconditional_guidance_scale 1.0 \
+        --ddim_steps 50 \
+        --ddim_eta 1.0 \
+        --prompt_dir "unitree_g1_pack_camera/case1/world_model_interaction_prompts" \
+        --dataset ${dataset} \
+        --video_length 16 \
+        --frame_stride 6 \
+        --n_action_steps 16 \
+        --exe_steps 16 \
+        --n_iter 11 \
+        --timestep_spacing 'uniform_trailing' \
+        --guidance_rescale 0.7 \
+        --perframe_ae \
+        --analysis_log_metrics \
+        --analysis_reference_steps 50 \
+        --head_schedule_steps 0 7 14 21 28 35 42 49 \
+        --head_skip_mode reuse_prediction \
+        --head_log_steps 40 43 46 47 48 49
+} 2>&1 | tee "${res_dir}/output.log"

unitree_g1_pack_camera/case1/world_model_interaction_prompts/unitree_g1_pack_camera.csv

@@ -0,0 +1,2 @@
+videoid,contentUrl,duration,data_dir,instruction,dynamic_confidence,dynamic_wording,dynamic_source_category,embodiment,fps
+0,x,x,unitree_g1_pack_camera,mount camera,x,x,x,G1_Dex1,30

unitree_g1_pack_camera/case2/output/inference/analysis/analysis_report.md

@@ -0,0 +1,54 @@
+# DDIM Analysis Report
+
+## Analysis Config
+
+- Stability threshold: None
+- Stability window: 3
+- Mid-step window: 15-35
+
+## Dataset Overview
+
+- Unique samples: 1
+- Step rows: 1100
+- Sample rows: 2
+- Round rows: 11
+- Scenes: unitree_g1_pack_camera
+- Pass types: policy, world_model
+
+## Five Directions
+
+### Original Early Stop
+
+- `latent_first_stable_step_policy`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_first_stable_step_world_model`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_l2_vs_full50`: mean=118.1448, median=137.2515, p90=165.2197, p95=172.7743
+
+### Direction C: Action Converges First
+
+- `action_first_stable_step`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_first_stable_step`: mean=nan, median=nan, p90=nan, p95=nan
+- `action_vs_full50_95pct_step`: mean=46.7273, median=46.7273, p90=46.7273, p95=46.7273
+- `share_action_first_stable_before_latent`: nan
+- `share_action_95pct_before_latent`: nan
+
+### Direction D: Cross-Step Similarity
+
+- `latent_delta_mid_steps`: mean=0.3529, median=0.3489, p90=0.4169, p95=0.4276
+- `action_delta_mid_steps`: mean=0.0313, median=0.0298, p90=0.0451, p95=0.0483
+- `state_delta_mid_steps`: mean=0.0310, median=0.0300, p90=0.0436, p95=0.0460
+
+### Direction A: Budget Heterogeneity
+
+- `oracle_budget_action`: mean=46.7273, median=46.7273, p90=46.7273, p95=46.7273
+- `oracle_budget_state`: mean=47.0000, median=47.0000, p90=47.0000, p95=47.0000
+- `oracle_budget_action_by_scene`:
+  - `unitree_g1_pack_camera`: mean=46.7273, median=46.7273, std=0.0000
+
+### Direction B: Round Reuse
+
+- `latent_init_dist_to_prev_round`: mean=236.4804, median=243.4704, p90=257.5054, p95=261.9766
+- `action_drift_vs_prev_round`: mean=0.2621, median=0.2435, p90=0.4256, p95=0.5909
+- `round_total_time_s`: mean=151.4911, median=151.4212, p90=152.2260, p95=152.4231
+- `policy_pass_total_time_s`: mean=74.2250, median=74.3476, p90=75.0393, p95=75.0564
+- `world_model_pass_total_time_s`: mean=74.2693, median=74.3542, p90=74.6399, p95=74.6751
+

unitree_g1_pack_camera/case2/output/inference/analysis/scene_summary.csv

@@ -0,0 +1,4 @@
+scene,pass_type,pass_total_time_s,pass_total_time_s,pass_total_time_s,action_first_stable_step,action_first_stable_step,action_first_stable_step,state_first_stable_step,state_first_stable_step,state_first_stable_step,latent_first_stable_step,latent_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_90pct_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,action_vs_full50_95pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_action,oracle_budget_action,oracle_budget_state,oracle_budget_state,oracle_budget_state,round_total_time_s,round_total_time_s,round_total_time_s,policy_pass_total_time_s,policy_pass_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,world_model_pass_total_time_s,world_model_pass_total_time_s,latent_init_dist_to_prev_round,latent_init_dist_to_prev_round,latent_init_dist_to_prev_round,action_drift_vs_prev_round,action_drift_vs_prev_round,action_drift_vs_prev_round
+,,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std
+unitree_g1_pack_camera,policy,74.22499192844738,74.22499192844738,,,,,,,,,,,44.90909090909091,44.90909090909091,,46.72727272727273,46.72727272727273,,46.72727272727273,46.72727272727273,,47.09090909090909,47.09090909090909,,151.4910772713748,151.4910772713748,,74.22499192844738,74.22499192844738,,74.26925277709961,74.26925277709961,,236.48041229248048,236.48041229248048,,0.2620766624808311,0.2620766624808311,
+unitree_g1_pack_camera,world_model,74.26925277709961,74.26925277709961,,,,,,,,,,,45.09090909090909,45.09090909090909,,46.81818181818182,46.81818181818182,,46.81818181818182,46.81818181818182,,47.0,47.0,,151.4910772713748,151.4910772713748,,74.22499192844738,74.22499192844738,,74.26925277709961,74.26925277709961,,236.48041229248048,236.48041229248048,,0.2620766624808311,0.2620766624808311,

unitree_g1_pack_camera/case2/output/inference/analysis/step_aggregate.csv

@@ -0,0 +1,102 @@
+pass_type,step,step_time_s,step_time_s,step_time_s,latent_delta,latent_delta,latent_delta,action_delta,action_delta,action_delta,state_delta,state_delta,state_delta,action_cosine_vs_full50,action_cosine_vs_full50,action_cosine_vs_full50,state_cosine_vs_full50,state_cosine_vs_full50,state_cosine_vs_full50,latent_l2_vs_full50,latent_l2_vs_full50,latent_l2_vs_full50
+,,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std
+policy,1,1.4643647887490012,1.4685146808624268,0.11946496958360509,1.4148771329359575,1.4147257804870603,0.0017984101608485988,0.04281757399439806,0.0392014160752296,0.008532584867967428,0.03407446464354336,0.0336839072406291,0.005709339832039931,0.08155293779617001,0.077914148569107,0.04679852520684907,0.07369075512344185,0.0933109745383262,0.06745704605608141,170.09748008034447,171.06507873535156,10.505909113191718
+policy,2,1.4541257944974033,1.4734899997711182,0.058318946913729415,1.0217632922259243,1.0221184492111206,0.0018243984267821882,0.03183789974586525,0.0312561765313148,0.0040836650740390815,0.02485884675248098,0.0250939335674047,0.0034012227624572343,0.08140638741579918,0.0768530443310737,0.04697127046684661,0.07446798606013708,0.0949154943227768,0.06698259561179119,169.23872514204547,168.94827270507812,10.557645930205437
+policy,3,1.4691599932583896,1.470665693283081,0.005543378191498716,0.850593398917805,0.8505691289901733,0.0010002504677263727,0.02459242144091556,0.0238988660275936,0.005148060909606215,0.018639818849888655,0.0189135950058698,0.0026148793321467243,0.08384929055517366,0.0789375901222229,0.0470194288131813,0.07829395512288263,0.0986015349626541,0.06741940188036281,169.08106717196378,170.78274536132812,10.922710014605762
+policy,4,1.4598333835601807,1.4700920581817627,0.03999101236346604,0.7504921176216819,0.7511603236198425,0.002204449431442244,0.01972710713744159,0.0199761055409908,0.00393638778985888,0.0155389210717244,0.0144460797309875,0.0029508824613728476,0.0880777879872105,0.0838154405355453,0.04698446571781875,0.0843947429887273,0.1043250784277916,0.06791345800638893,168.03258167613637,169.22280883789062,10.922582818026411
+policy,5,1.4335628856312146,1.4701387882232666,0.0808474674398391,0.6826300404288552,0.6819959282875061,0.001890172866151874,0.01639852198687462,0.0156712140887975,0.002790387852083108,0.014331226067786819,0.0129353087395429,0.0034085937748645143,0.0936777947301214,0.0910099148750305,0.04711223904805821,0.09181853510778057,0.1118631586432457,0.06836859573484567,166.0841203169389,166.22369384765625,10.243451847158397
+policy,6,1.4402985572814941,1.4712324142456057,0.06835911239779546,0.6330406774174083,0.6331149935722351,0.0013513106855039577,0.0151571499353105,0.0149793233722448,0.0031206323648545163,0.012832071208818345,0.0122287198901176,0.0023915636981314847,0.10016470639543094,0.0993154644966125,0.04738779037041065,0.10003195347433737,0.1204004138708114,0.06860743916320584,165.64228959517047,165.88099670410156,10.263589696848832
+policy,7,1.451158111745661,1.4723591804504397,0.044411611953512946,0.594232824715701,0.593671977519989,0.0011183720417689005,0.015228892591866544,0.014497703872621,0.003513930042467193,0.012337777763605064,0.012337090447545,0.001362647107553788,0.1072305437516082,0.1087293922901153,0.04779812449365048,0.10901441539383744,0.1298665702342987,0.06866700632653792,162.61344216086647,162.01109313964844,9.965445014416865
+policy,8,1.453081651167436,1.4714808464050293,0.060856995323781735,0.5634365190159191,0.5633569359779358,0.001538824837380747,0.015222255140542927,0.0144889457151293,0.0036440824929860165,0.012728041545911218,0.0126567129045724,0.0011071913735304527,0.11505048755895002,0.1191661655902862,0.04828471178990196,0.11853112077171148,0.1399197578430175,0.06863627871831689,160.31207552823153,159.6591796875,9.998474862005931
+policy,9,1.4518233862790195,1.4713809490203855,0.06384766825417784,0.5371716997840188,0.5367397665977478,0.00119799520290687,0.015034105459397427,0.0141661390662193,0.0031497299872971377,0.013244852677664927,0.0129782455042004,0.0012503706354094228,0.12357755276289849,0.1302836835384369,0.04875568816050281,0.1285001942887902,0.1504424214363098,0.06854927892680253,158.74315990101206,158.91424560546875,9.334594090177877
+policy,10,1.45024620402943,1.4693543910980225,0.04926523061596906,0.5157278776168823,0.5156438946723938,0.0006784788635188598,0.015083106234669656,0.0145043134689331,0.0025643355262319,0.013690607198937327,0.0135310553014278,0.0013257226314623452,0.1328831558877771,0.1421728134155273,0.048996699101949974,0.1388687498190186,0.1613283008337021,0.06845518924164967,156.9241804643111,159.1904754638672,8.110437410818491
+policy,11,1.444454388184981,1.4709253311157229,0.0885492632473059,0.496602090922269,0.4962410032749176,0.0013039866938400913,0.015376358716325273,0.0151475789025425,0.002017205620873061,0.014698531732640428,0.0144659560173749,0.0013280586819555027,0.14287893541834568,0.1525715142488479,0.04904092680134504,0.14972653985023496,0.1728588342666626,0.06843626264144984,155.5454961603338,157.67005920410156,7.781809460759908
+policy,12,1.4495672095905652,1.4743444919586182,0.08235753379631829,0.4801749966361306,0.4799525737762451,0.000902742483838771,0.015821124935014638,0.0156387612223625,0.0017984735074255864,0.0155477133494886,0.0153720257803797,0.0012135653508671457,0.15324927392331034,0.1628736108541488,0.04897643937871421,0.16094945540482344,0.184751346707344,0.0684884090749342,154.19762906161222,156.07546997070312,8.150938548918997
+policy,13,1.4486402164806018,1.4744722843170166,0.07969306661445948,0.46393065831877967,0.4638084769248962,0.0008748368673454794,0.016626007177612954,0.01647275313735,0.0018086679872989675,0.016100792298940046,0.0161943007260561,0.0012045572244640211,0.16416988725011994,0.1732068806886673,0.04896005116781644,0.1725476736372167,0.1969069391489029,0.068478736709341,152.0845323042436,153.36557006835938,7.721437583934334
+policy,14,1.4453192190690474,1.468017339706421,0.05264109303843239,0.4508701535788449,0.4505763351917267,0.0009505103780250265,0.017609759657220354,0.0175512973219156,0.0017592237890723748,0.017064232294532328,0.0170348230749368,0.001270485708400615,0.17561168088154355,0.1836914122104644,0.048965035477324795,0.18447633697227994,0.2095178365707397,0.06836240386069921,150.21925631436434,151.2340545654297,7.28035098163076
+policy,15,1.4479859308762983,1.4688806533813477,0.07842091803812619,0.43892092596400867,0.4388017356395721,0.0006090699321633986,0.018392478708516416,0.018379459157586,0.0016634864046166824,0.0178374705666845,0.0178643409162759,0.0013032702419437212,0.18761804970827967,0.1929938793182373,0.048950300426178504,0.19682544266635718,0.2224430739879608,0.0681496771901637,148.9786931818182,149.8619842529297,7.183796803551464
+policy,16,1.4501507932489568,1.472064971923828,0.07296202797146564,0.4266620549288663,0.4265547692775726,0.0006317364362079431,0.019098526713522943,0.0191935021430253,0.0014391345328292882,0.018909056586298028,0.0189465321600437,0.0012669991546541315,0.20013522695411334,0.2009296715259552,0.04894724949745174,0.20969889245250004,0.235941857099533,0.06796722492792714,147.64975807883522,148.93096923828125,7.028528784118946
+policy,17,1.4665821682323108,1.466827392578125,0.010499568350019747,0.41590515591881494,0.416337251663208,0.0009997707025206652,0.019994667646559754,0.0201997961848974,0.001379925491732909,0.019927606325257856,0.0198932811617851,0.001317283947292341,0.2131885994564403,0.2105266749858856,0.04890426258779776,0.22302113812078125,0.2498722970485687,0.06779833250998542,146.45462868430397,148.46408081054688,6.73497239911813
+policy,18,1.4712181308052756,1.474581956863403,0.007750363440223481,0.40625025738369336,0.4059191048145294,0.0010800856216576027,0.020714854652231336,0.0208488684147596,0.001384568576282785,0.02073309655216601,0.0206792298704385,0.0012687780651277735,0.2267513667995279,0.2258680015802383,0.048743387281005676,0.23679225214503025,0.2641832828521728,0.06771691137615982,145.28262329101562,147.085693359375,6.443550109193905
+policy,19,1.4439404877749356,1.4710924625396729,0.06232354674563371,0.3967506912621585,0.396649420261383,0.0010050803948196772,0.022012898698449093,0.0223398562520742,0.0012621234090054861,0.021984958851879257,0.0218807458877563,0.001258675344227013,0.24069879542697556,0.2418799698352813,0.04880449050600198,0.25113794884898444,0.2790946960449219,0.06758273720249028,144.20575228604403,146.3004913330078,5.954482749653211
+policy,20,1.4693304191936145,1.4748013019561768,0.011052732221516794,0.38752706484361127,0.3874090313911438,0.001034296580485665,0.022933603518388454,0.0232044085860252,0.0013018049255700318,0.0228283342651345,0.0226724166423082,0.0012637530791815457,0.25539551539854566,0.2567644119262695,0.04870047432505133,0.2658977820114656,0.2942913472652435,0.06748155620320793,142.94178355823863,144.65557861328125,5.462257314486613
+policy,21,1.4739628488367253,1.475623607635498,0.004919572374213548,0.378894952210513,0.3791788518428802,0.001025993760118785,0.024044604125347955,0.0242867935448884,0.00132583235865487,0.024273770268667772,0.0241147838532924,0.0012500849623744918,0.2706766602667895,0.2709282636642456,0.048556081192350074,0.2812915945594961,0.3102010488510132,0.06739307856150614,141.88197049227628,143.0328369140625,5.328747554157212
+policy,22,1.4712780605662952,1.4691743850708008,0.006864379727599383,0.3710618181662126,0.3712555766105652,0.0006692269219455867,0.025322464379397266,0.0255931634455919,0.0012715289211763724,0.02550048990683117,0.0252727624028921,0.0012870320310288282,0.28641768883575086,0.2852920591831207,0.04842803930383065,0.2973022406751459,0.3266588449478149,0.06725796773315261,140.9632929021662,142.11849975585938,5.0136914538419495
+policy,23,1.4350546273318203,1.4687039852142334,0.07997943289279814,0.3633401122960177,0.3631722331047058,0.0009989748146953473,0.026672355153343843,0.027136480435729,0.0012443973114825869,0.026658307083628353,0.0263877399265766,0.0013472910980284953,0.3025903904979879,0.3030104041099548,0.048280787305788025,0.3139006414196708,0.3423101007938385,0.06703547592704566,140.0803319757635,140.9939422607422,4.593726109044367
+policy,24,1.4309492761438543,1.4715476036071775,0.0858298076934187,0.3563299693844535,0.3563564717769623,0.0008959700858330447,0.02793866802345618,0.0282828044146299,0.0014388407803073353,0.02799780582162463,0.0276559609919786,0.0014552222541383201,0.3197950368577783,0.3229982256889343,0.04783274846005909,0.33120391450145026,0.3564364314079284,0.06672903498922225,139.2905828302557,140.40904235839844,4.435610070156716
+policy,25,1.4494194117459385,1.4702398777008057,0.07093145311155574,0.3493931158022447,0.3492731750011444,0.0009851462446976535,0.029285558414730108,0.0295346230268478,0.001505166961002917,0.02943666093051428,0.0291007291525602,0.0015311789550177448,0.33794105865738605,0.3434889912605285,0.047160371525491775,0.349244935946031,0.3712327480316162,0.06638113111553619,138.36009493741122,139.26083374023438,4.1431081850152776
+policy,26,1.4699790261008523,1.4730641841888428,0.01345044089820449,0.34222279082645063,0.3419949114322662,0.0010486632211865125,0.030812483111565717,0.0313926599919796,0.001548341935311138,0.0306619200855493,0.0303577240556478,0.0015910381564822266,0.35674967007203534,0.3649344444274902,0.046474867806131706,0.3679815882986242,0.3865175247192383,0.06595508811603952,137.8521201393821,138.7310028076172,4.128792787821338
+policy,27,1.468766754323786,1.4721136093139648,0.011633283159055248,0.33555776422674,0.335639089345932,0.001213625565416701,0.032442700287157795,0.0328396186232566,0.001562470976658336,0.03222530335187906,0.031873095780611,0.0016364673207481441,0.3764675611799413,0.3874861896038055,0.045751812835412776,0.38755452091043646,0.4025210440158844,0.06542703424857234,137.35802806507456,138.27166748046875,3.91192755951621
+policy,28,1.453011079268022,1.4696829319000244,0.05923218778920788,0.32929666746746405,0.329518586397171,0.0011773860605221414,0.03402061015367503,0.0343900397419929,0.0016623573205540526,0.033716790716756435,0.0334406234323978,0.0016943412683561483,0.3970481888814406,0.4109510481357574,0.04481957298571838,0.40797183459455316,0.4191781878471374,0.06482322389922324,136.80251797762784,137.7513427734375,3.842427618704297
+policy,29,1.4546096108176492,1.4728519916534424,0.06608759137741625,0.3229366459629752,0.322741150856018,0.0012239891389898506,0.0356819778680801,0.0359496474266052,0.0019054730060930312,0.03533605147491797,0.0348477996885776,0.001774232015556231,0.4189335730942813,0.4347765743732452,0.04373257254016122,0.42934841188517486,0.4366675019264221,0.06408464580588284,136.07441295276988,137.02008056640625,3.616032880010223
+policy,30,1.4577914151278408,1.474189043045044,0.053860776170227574,0.31670473109592084,0.3166425228118896,0.0011014849973054641,0.03759356588125225,0.037754151970148,0.0018783720256698863,0.037365547974001226,0.036752238869667,0.0019108722852618706,0.441381879828193,0.4540104866027832,0.04289638972643966,0.45185111327604816,0.4552650451660156,0.0631586537050016,135.57533541592684,136.5150146484375,3.4262039482242024
+policy,31,1.4538111253218218,1.4696011543273926,0.0360097670839535,0.31162824143062934,0.3117263913154602,0.0009013153305599144,0.0393330918794328,0.0392529480159282,0.0019819057134478056,0.039152730256319,0.0384356975555419,0.001969918607737662,0.4651488878510215,0.4741113483905792,0.041995091315069094,0.47539693117141724,0.4748137593269348,0.062114575384912496,129.24837285822088,129.9188232421875,3.2632438024444537
+policy,32,1.4704401926560835,1.474531650543213,0.01075551266362713,0.30393185398795386,0.3042071461677551,0.0010586996264491486,0.04157216542146417,0.0412184819579124,0.0017972823761834788,0.04045209728858683,0.0396092012524604,0.002139237437079005,0.4897051291032271,0.4956349432468414,0.04134991595638768,0.4998013106259433,0.4950272142887115,0.06090836017505551,122.95333307439631,123.65977478027344,3.0880454472713708
+policy,33,1.4731678745963357,1.473830223083496,0.004752219302401104,0.2957445383071899,0.2958968877792358,0.0011429464467950909,0.04347564313899382,0.0428317114710807,0.0024512506577975828,0.042609430510889355,0.0415723286569118,0.002255994799632196,0.5162798681042411,0.5207837224006653,0.04007511057883185,0.5254944102330641,0.5165548324584961,0.05949591865922783,116.75520671497692,117.48394012451172,2.9589432387265497
+policy,34,1.4589084061709316,1.4665017127990725,0.032630350812439486,0.2876056080514734,0.2876955866813659,0.0013782416113859587,0.04568399996920061,0.0449166558682918,0.002378854999547093,0.04469157010316844,0.0436306819319725,0.0023669127989235706,0.5440903929146853,0.5513156652450562,0.03890638015384111,0.5524050755934282,0.5393847227096558,0.05786256046399864,110.42148936878552,111.21601104736328,2.838917270477883
+policy,35,1.4514379718086936,1.4714417457580566,0.04617745351386486,0.27884076129306445,0.2791093587875366,0.001043473372444849,0.04801188646392384,0.0471608825027942,0.002393580620760842,0.04675462469458577,0.0458813868463039,0.0025468147949929054,0.5733803022991527,0.581934928894043,0.037760664059080656,0.5805426077409224,0.568490743637085,0.056002218845501696,104.15709408846769,104.91960906982422,2.7979038383335135
+policy,36,1.4492437189275569,1.471243143081665,0.0737765238865744,0.27016199718822126,0.2702356874942779,0.0011306823207609445,0.050716660916805226,0.0497601516544818,0.002625188680908781,0.04999401047825809,0.0494257248938083,0.002795705750731449,0.6042665568265048,0.6092218160629272,0.03640729869526877,0.6104725436730818,0.6012518405914307,0.05381268756361428,97.95745918967508,98.840576171875,2.738723778717534
+policy,37,1.4538352489471436,1.4690988063812256,0.0457836322023699,0.26096625219691877,0.2610688805580139,0.0009061107546286581,0.053497848524288634,0.0527910850942134,0.003212518277271837,0.051143160597844496,0.0507018715143203,0.0029951523969799073,0.637215646830472,0.6398462653160095,0.03444890247036076,0.6409264532002535,0.6355582475662231,0.051507657595687664,91.652099609375,92.54032135009766,2.684004496466761
+policy,38,1.4566808180375532,1.4724955558776855,0.04235987573073707,0.2516751993786205,0.251907616853714,0.001117988117132102,0.05602727424014693,0.0561307556927204,0.00328600563875306,0.05494774200699542,0.0549914576113224,0.00332624741891855,0.6713680679147894,0.6724480390548706,0.03229591071369006,0.6733194589614868,0.671265721321106,0.0487047944466783,85.39000909978694,86.12053680419922,2.627467613788882
+policy,39,1.4406840801239014,1.47485613822937,0.08354631830102618,0.2420696534893729,0.2423757165670395,0.0012514525165951433,0.05881078338081183,0.059526477009058,0.003418186954718648,0.057084265419027974,0.0575138628482818,0.003645173096403443,0.7066746191544966,0.7066949009895325,0.030047146565527597,0.7065333344719626,0.7081829309463501,0.045673785779410815,79.27400415593928,79.92479705810547,2.5960672292479985
+policy,40,1.4064905426718972,1.4690375328063965,0.11585307183135446,0.23244414952668271,0.2326675355434417,0.0013495055124272922,0.06157916411757465,0.062261588871479,0.003564051416678765,0.05833273956721474,0.059008415788412,0.0038470968256900068,0.7427707747979597,0.7417453527450562,0.02763995074485661,0.739755776795474,0.7450113296508789,0.04246605478348856,73.0922588001598,73.5637435913086,2.4314005693356804
+policy,41,1.4313715371218594,1.466968297958374,0.07368545861305165,0.22199878096580503,0.2217127829790115,0.0013969031394365553,0.06610668315128843,0.0665889084339141,0.0035829755370485484,0.06246782974763346,0.0629662051796913,0.004400222545127881,0.7797025496309454,0.7775286436080933,0.0251589907318902,0.7743967554785989,0.7825913429260254,0.03867995266953225,66.98066642067649,67.30145263671875,2.327367054318163
+policy,42,1.428111726587469,1.4738636016845703,0.10034509469545733,0.21233072064139621,0.2122133672237396,0.0014393924320923032,0.06758632578633042,0.0684843808412551,0.0038116478602509816,0.06470932709899811,0.064999796450138,0.0047624362430992205,0.8161974115805193,0.8139770030975342,0.022367948765810382,0.8088300011374734,0.8194238543510437,0.03460052082312121,60.793091167103164,61.25485610961914,2.301073842393202
+policy,43,1.4665637883273037,1.4747726917266846,0.021786420640368755,0.20184920186346222,0.2016905844211578,0.0016286538498986188,0.07008000598712397,0.0715215727686882,0.004076197175567837,0.06775860082019455,0.067810982465744,0.005260289591197391,0.8518163182518699,0.8503491282463074,0.01934837710554393,0.8429664210839705,0.8516335487365723,0.030134335252600654,54.72544756802645,55.17553329467773,2.206402025876178
+policy,44,1.4727906313809482,1.4726550579071045,0.004036726084646233,0.19168418374928556,0.1914531886577606,0.00140539550225344,0.07191436195915392,0.0736697241663932,0.00439316234216171,0.07105440849607637,0.0708177387714386,0.005900156371618375,0.885437851602381,0.8852209448814392,0.016208972434463058,0.8762508739124645,0.8820749521255493,0.02520525922544915,48.51811391657049,49.013309478759766,2.1088416380569814
+policy,45,1.446694547479803,1.4745047092437744,0.07551140740072294,0.1812714406035163,0.181238442659378,0.0014373250432260674,0.0742860836061564,0.0762761756777763,0.004684855683469112,0.07260908457365899,0.0721836537122726,0.006349689097127822,0.9164211587472395,0.9174055457115172,0.012989724729321951,0.9070658846334978,0.9110933542251588,0.020215141180785,42.32746817848899,42.7825813293457,1.9662650589241975
+policy,46,1.4368385401639072,1.4654018878936768,0.08385292373748879,0.17085370556874704,0.1709097772836685,0.0014258611868877991,0.07530549033121627,0.0774233341217041,0.005084526833169449,0.07631432739171112,0.0748278275132179,0.007228668594895186,0.9434121738780629,0.9452750086784364,0.009739925980528348,0.9354661919853904,0.9390493035316468,0.01502288706537057,36.078021656383164,36.60586166381836,1.8736266830423625
+policy,47,1.4586541652679443,1.4740231037139893,0.052442128026293486,0.16035677357153455,0.1603856384754181,0.0014218573660375136,0.07815956798466765,0.079449750483036,0.005649116803440019,0.08074588599530129,0.0784367695450782,0.008250123508504568,0.96619666706432,0.968171000480652,0.006511558222465456,0.9604470025409352,0.9631124138832092,0.009887607373441406,29.675485957752574,30.196746826171875,1.710546362843461
+policy,48,1.4722720926458186,1.4723389148712158,0.006367569543143773,0.14940655908801334,0.1492402106523513,0.0013993454651778864,0.08010110394521187,0.0802122130990028,0.0065067700622938075,0.08444487845355811,0.0823443830013275,0.009396625209948162,0.9836481647057966,0.9846829771995544,0.003527035088416307,0.9803741194985129,0.981927752494812,0.0053060435890045525,22.973013444380328,23.38234519958496,1.414627103345284
+policy,49,1.444547176361084,1.4667015075683594,0.07979592155859828,0.1370534097606485,0.1371117383241653,0.0016530970261065567,0.0824336091225797,0.0813702046871185,0.007697837816589516,0.0850304785099896,0.0825256034731864,0.010166840989871232,0.9950967051766135,0.9953904151916504,0.0011709772222340735,0.9935735464096069,0.9941732883453368,0.001927078989020655,15.557310711253773,15.786161422729492,0.9877801119379396
+policy,50,1.4521542245691472,1.4720304012298584,0.06603852925661893,0.11369569125500588,0.1138272359967231,0.0023110222589311866,0.10032822530377991,0.100488431751728,0.010894687851971983,0.11266036467118692,0.1085293665528297,0.016299536150999543,1.0,1.0,6.529361694285094e-08,1.0000000108372082,1.0,1.2162322090147332e-07,0.0,0.0,0.0
+world_model,1,1.4531300067901611,1.4690628051757812,0.0491156866948184,1.4132896228270098,1.4142149686813354,0.003255941925227477,0.04233239455656567,0.0411906465888023,0.005088275729587036,0.02972275696017521,0.0294788870960474,0.002841034427781072,0.01066405207596041,0.0190485790371894,0.09199178404776863,0.02096229859373785,0.0179943591356277,0.05441100021739327,170.52376487038353,170.8175811767578,11.959971483244104
+world_model,2,1.445748134092851,1.468601942062378,0.08171767080670449,1.0209344842217185,1.021303653717041,0.0015422980053260021,0.03186475400897584,0.0320959351956844,0.003721582363586933,0.022558428685773454,0.0224028918892145,0.0018216906440364554,0.008729932829737663,0.0177913960069417,0.09297319810205869,0.021638473834503763,0.0198191553354263,0.05502681677640681,169.82606506347656,169.7007598876953,11.959937134835702
+world_model,3,1.4506308165463535,1.4671649932861328,0.060800746781730586,0.8507153229279951,0.8504046201705933,0.0018239936759067248,0.0240144531496546,0.0244467742741107,0.003459577360881089,0.017316160554235592,0.0169892609119415,0.0011381811987599272,0.010169191963293328,0.0220091436058282,0.0937136821672725,0.025857549160718904,0.0239264145493507,0.05581771555239258,169.58072731711647,168.5248260498047,12.44009009157486
+world_model,4,1.4488683180375532,1.4695987701416016,0.06083445703556617,0.7501633275638927,0.7506617307662964,0.0019104562769016382,0.019725087098777235,0.0205037556588649,0.0029169759726555044,0.014891516332599199,0.0146292811259627,0.0014052944916993554,0.013635958798907007,0.0289527475833892,0.09436679755241259,0.032324665903367765,0.0300790797919034,0.05650057724582583,169.13958740234375,167.8528594970703,12.606348203149023
+world_model,5,1.4711864211342551,1.4697744846343994,0.005543605858954345,0.6832541877573187,0.6833713054656982,0.001706121726052321,0.016979324038733073,0.0177939422428607,0.0020666375628348465,0.014220128428529545,0.0139124719426035,0.0022124306807632956,0.018689718774773864,0.0375685915350914,0.09471027705698441,0.040026255586946544,0.0374040566384792,0.05707607130608291,167.6177714954723,166.66705322265625,11.796981533614542
+world_model,6,1.469280112873424,1.4719183444976809,0.007035133789255347,0.6333861784501509,0.6336749792098999,0.0021994267898777678,0.015492973087186118,0.0159195940941572,0.0019235300148187366,0.014116137288510737,0.0133379511535167,0.0031382793140078396,0.024794587188146335,0.0456058643758297,0.09487748563846209,0.04845040494745425,0.0457378253340721,0.05747753645711032,165.6778994473544,166.3711395263672,10.619013098115387
+world_model,7,1.472085866061124,1.4723243713378906,0.006705820056056248,0.594574744051153,0.5951471328735352,0.0017231972182991469,0.015496570870957545,0.0154151702299714,0.0018451227369357193,0.014593197371471972,0.0136789213865995,0.003112130334601752,0.03163379735567353,0.0531187243759632,0.09500268267607594,0.05747796129435297,0.054626677185297,0.05784256562113932,164.5299432927912,167.68191528320312,10.895195708066765
+world_model,8,1.4720198457891291,1.4723889827728271,0.00518391152535472,0.5636271129954945,0.5632905960083008,0.0017041149849807641,0.015425535088235636,0.0149150779470801,0.002022114955101204,0.015047084963457101,0.0144340507686138,0.002766965369977585,0.039239831099455995,0.0612411201000213,0.09529415723945427,0.06695743971927595,0.0640106722712516,0.05829367138748683,161.57355013760653,163.56802368164062,9.20971537732733
+world_model,9,1.469901821830056,1.4702527523040771,0.004054648873007827,0.5375342098149386,0.5376063585281372,0.0022408520244777924,0.01536705979908049,0.0150943789631128,0.0020627140424513473,0.015495549379424566,0.0150679098442196,0.0025499023501039187,0.047641504047946495,0.0695167630910873,0.09545233680869104,0.07685744305225933,0.0736455470323562,0.05871498975218288,159.50396451083097,162.58323669433594,8.806904337789945
+world_model,10,1.4392199516296387,1.4625244140625,0.08112541237824192,0.5162806239995089,0.5160682201385498,0.001015326527604176,0.01581603263250801,0.0156884975731372,0.002194845011028816,0.015158226459541053,0.0151408966630697,0.001827411956165913,0.05681518075818364,0.0773289203643798,0.09557560321661594,0.08726916614581236,0.0845247060060501,0.0590524145616215,157.6842609752308,158.38314819335938,7.924084492832992
+world_model,11,1.458111589605158,1.4677298069000244,0.043944208682374794,0.49642668528990314,0.496480256319046,0.0005663056442703679,0.016232124953107337,0.0161107461899518,0.00202627913709701,0.016051699746061417,0.0156848207116127,0.0023461394014470966,0.0667126461524855,0.08636424690485,0.0957183652896705,0.0982694817198948,0.0955507382750511,0.059556103353279065,155.52410195090553,156.44131469726562,7.5939063999666
+world_model,12,1.4709015976298938,1.4696195125579834,0.006571165767571775,0.4789586663246155,0.4792364537715912,0.0013763685255176977,0.01668058818375519,0.0165126323699951,0.002024669784464341,0.01710942666977639,0.0166090708225965,0.0025077171602272,0.07705980844118376,0.095755249261856,0.09579551584430755,0.1097479498183185,0.1067567467689514,0.06021688248046881,152.93443991921166,155.23826599121094,7.2050821136123355
+world_model,13,1.4617737206545742,1.4697272777557373,0.023172067070014075,0.463858271187002,0.4638905525207519,0.0009477019034480374,0.017425104213709134,0.0167829282581806,0.0022162356934673907,0.0171601602976972,0.0172439478337764,0.0012111169653336873,0.08788447149775241,0.1056859940290451,0.09596174211067532,0.12172558565031394,0.1184326708316803,0.06080954837990382,151.11357532848012,153.3643341064453,6.853307435947781
+world_model,14,1.454773035916415,1.4701380729675293,0.03978065430192767,0.4507119493051009,0.4507866203784942,0.0011219127697888148,0.018485929139635728,0.0187662951648235,0.002262044974358526,0.01810785349119788,0.0182711947709321,0.001235132468826602,0.09915354492312126,0.1164980083703994,0.09619746304615621,0.134206107055599,0.1303138434886932,0.06138013568876015,150.10934448242188,150.59906005859375,6.446610288789346
+world_model,15,1.4720990657806396,1.469578981399536,0.007278597676559286,0.4381364177573811,0.4380443394184112,0.0009090200226474513,0.01926086301153351,0.0200432352721691,0.0019577541730104123,0.01886712945997711,0.0188689120113849,0.0013236201255172066,0.11103096494281829,0.1281785815954208,0.09631354776925964,0.147093788805333,0.1426986157894134,0.06198671636642755,149.1476260098544,151.31027221679688,6.635135587715655
+world_model,16,1.4700754555788906,1.471893072128296,0.007316474803078249,0.4262196123600006,0.4262261390686035,0.001095853726906011,0.01975194686515758,0.0202637277543544,0.0015447396570022426,0.02004082263870667,0.0198499336838722,0.0014795669134745663,0.1235810842111029,0.1403981745243072,0.09632255942860052,0.16050492972135538,0.1562183201313018,0.0626203087188999,147.29337241432884,148.6558074951172,6.348471381299838
+world_model,17,1.466769196770408,1.4713497161865234,0.01920632783168187,0.4154950895092704,0.4158066213130951,0.0008478264439112066,0.020501015369187654,0.0203900691121816,0.0014808879254982615,0.02109600095586337,0.0207161176949739,0.0014869817054661667,0.1367613773115656,0.1531966775655746,0.09620123742728133,0.17452066662636667,0.1704875230789184,0.06324886632959478,145.88161399147728,147.55361938476562,5.950619793837957
+world_model,18,1.4245793169195002,1.466609239578247,0.08345920226518164,0.40545678680593317,0.4053582847118377,0.0011237834337533212,0.02135313386944204,0.021364402025938,0.001510699635536334,0.02179946547204792,0.0214679613709449,0.0012503401103612277,0.1505665535276586,0.1667198240756988,0.09589316049999586,0.1890738877383145,0.1857475638389587,0.06381919235034304,144.5868058638139,145.1148223876953,5.7265322391357065
+world_model,19,1.423682082783092,1.4692938327789309,0.09700786789595284,0.3961623961275274,0.3964757025241852,0.000765055684079319,0.02290288934653452,0.0230877380818128,0.0015306211435626655,0.02287882075391028,0.0228860843926668,0.0011773512016295274,0.16465594450181179,0.1806427240371704,0.0957312200780837,0.20437517220323737,0.2016612589359283,0.06429525441824575,143.37724581631747,143.6531982421875,5.1122679057288165
+world_model,20,1.4519510485909202,1.470816135406494,0.06540100213899629,0.3870067596435547,0.3869202136993408,0.000961959639389731,0.023834095421162473,0.0237349737435579,0.0015648372774805677,0.023748547177423064,0.0238281153142452,0.0012685565419330582,0.17962736162272364,0.1953206956386566,0.095270043596291,0.22021850808100263,0.2184108197689056,0.06463873858275421,142.57265403053978,142.62074279785156,4.8540946600338755
+world_model,21,1.469086386940696,1.4733264446258545,0.016609169191790514,0.3790786862373352,0.3790406584739685,0.0007978135176856313,0.025076422332362666,0.0248631089925766,0.0016299939810204798,0.025263442911884957,0.0254557244479656,0.0012033148992342787,0.195294512943788,0.2109169960021972,0.09469565248011942,0.23681880127299912,0.2354452311992645,0.064970520672511,141.58659501509234,141.7988739013672,4.6444953880450255
+world_model,22,1.4610871401700107,1.4718267917633057,0.028261280196480453,0.370963299816305,0.3706280291080475,0.0013548838603203359,0.02639219452711664,0.0262577179819345,0.0016122160471264543,0.02656346034597261,0.0267395377159118,0.001219294834088595,0.21144903518936847,0.2271564900875091,0.09418362144212566,0.254150237549435,0.2533199191093445,0.06525598351944148,140.5421142578125,140.31036376953125,4.3373264141730825
+world_model,23,1.430451826615767,1.467644453048706,0.09233157488571077,0.3632808138023723,0.3630765080451965,0.0010518236752857677,0.027740300379016126,0.0276239644736051,0.001692139154519931,0.027793135324662325,0.0279642567038536,0.0012787554955043343,0.22804807194254612,0.2436566352844238,0.09365754357826359,0.27216048403219745,0.2720200717449188,0.06547643188348523,139.64239224520597,139.7513885498047,4.0072054801565775
+world_model,24,1.4323539300398394,1.463529348373413,0.06201989034066444,0.35583845864642744,0.3557867705821991,0.0007232099504045768,0.02893048728054213,0.0290847942233085,0.0018424511186074359,0.029150741174817044,0.0294953100383281,0.001338676232192966,0.24603060362013898,0.2615217268466949,0.09276969892646586,0.2908760255033319,0.2909644246101379,0.06564597080834653,138.94943098588422,139.13819885253906,3.834489847496233
+world_model,25,1.4633264541625977,1.464641809463501,0.02652451469429186,0.3488414802334525,0.3485233783721924,0.0007312557415958559,0.030372981659390674,0.0305409003049135,0.0019272411184883826,0.030580269680781747,0.0309873539954423,0.0013684310172275187,0.2650862119414589,0.280240535736084,0.09169564359099053,0.3103855347091501,0.3107283115386963,0.06577745511964928,138.19119123979047,138.58604431152344,3.6427023787978094
+world_model,26,1.471280661496249,1.471991777420044,0.005683222740143182,0.34183617613532324,0.3418156802654266,0.0006849644418783603,0.03189678913490334,0.0321813002228736,0.0019194438896496214,0.031875828789039044,0.0322826467454433,0.0014988002778979373,0.2848215739835392,0.2996532320976257,0.09059124512454203,0.3306668983264403,0.3315136432647705,0.06581576767184823,137.6698455810547,137.97390747070312,3.5267321682102835
+world_model,27,1.4708820256319912,1.4704248905181885,0.006149145297447079,0.33516943454742426,0.3349823355674743,0.0011949821057581013,0.03362363678487861,0.0337959863245487,0.0020381003233469397,0.033479412678967764,0.033899713307619,0.0015586355252730606,0.30558137866583734,0.3200231790542602,0.08935057715614535,0.3518256626345894,0.3529286086559295,0.0657688654563084,137.11125599254262,137.46824645996094,3.3644789833366224
+world_model,28,1.441519173708829,1.4663872718811035,0.06422803048425728,0.329233705997467,0.3295179903507232,0.0007399095741990369,0.035295867615125334,0.0349985659122467,0.0022081246445453556,0.035003838214007245,0.0352316200733184,0.0016441857146907055,0.3273949135433543,0.3419556319713592,0.08793338409684347,0.3738611475987868,0.3748188018798828,0.06565447277865126,136.57470703125,136.7288360595703,3.1752900185615993
+world_model,29,1.4498719518834895,1.4688105583190918,0.05804785962166953,0.3227106712081215,0.3228993117809295,0.0010406739822097508,0.03706172989173365,0.036040112376213,0.00259055321401822,0.03660989078608421,0.0367550514638423,0.0017441106157484963,0.3506247414783998,0.3643530607223511,0.08609273737438285,0.3969181478023529,0.3978601098060608,0.06546277468117598,135.87558122114703,136.01229858398438,2.9250949953998204
+world_model,30,1.4540677937594326,1.4669373035430908,0.05436802292847809,0.3162551413882862,0.3162784278392792,0.0008230656428610575,0.038940347392450635,0.0384159609675407,0.0026545873186484066,0.038709479299458546,0.0390499643981456,0.0018845285748744277,0.3744500116868453,0.3832985460758209,0.08433074452507823,0.4212214106863195,0.4215447306632995,0.06510708647964135,135.42417214133522,135.47581481933594,2.843754566132659
+world_model,31,1.4611001014709473,1.4673335552215576,0.01873146363052879,0.3114192404530265,0.3114498555660248,0.0006467619653835257,0.04100383784283286,0.0404513664543628,0.00310486711823888,0.04051379893313751,0.0411303639411926,0.001995572749283914,0.40009593150832434,0.4045018553733825,0.08197590579675838,0.44663512435826386,0.4462751746177673,0.06460112840083017,129.15057581121272,129.26109313964844,2.758435442153553
+world_model,32,1.4507885629480535,1.4678740501403809,0.056892712878584994,0.30365204269235785,0.3037715256214142,0.000734843547571063,0.04331954636357043,0.0435155220329761,0.0029622720924499693,0.04180429774251846,0.0426932275295257,0.0021295857240458643,0.42660594257441437,0.4262927770614624,0.07998747179690448,0.47296674955974927,0.4756176769733429,0.0639694242969094,122.89028930664062,122.79805755615234,2.639768391084652
+world_model,33,1.441014441576871,1.4665913581848145,0.07428681475942614,0.2955560792576183,0.2959820926189422,0.001045356978288499,0.04558015072887589,0.0455912612378597,0.003786444172962228,0.0439923497086221,0.0450298748910427,0.0022740255547450453,0.4556803730401126,0.4582681953907013,0.0769823800534226,0.5006368458271027,0.5065355896949768,0.06313922632368967,116.62757179953836,116.40681457519533,2.6196776455877386
+world_model,34,1.4260302890430798,1.470289945602417,0.09875091569117307,0.28713381290435785,0.2875880897045135,0.0010804516266425688,0.04798809337345033,0.0487234182655811,0.0038678485454233017,0.04608547247268932,0.0471869967877864,0.002454442022208861,0.48608616536313837,0.4899641275405884,0.07390868023112801,0.5296258303252134,0.538323700428009,0.06205317851664674,110.36315155029297,110.11024475097656,2.5306731578704436
+world_model,35,1.4486090486699885,1.4718477725982666,0.06079356298040526,0.2787444158033891,0.279075026512146,0.000948876299882842,0.05060343105684622,0.0517271868884563,0.004066411831685735,0.04811720245263788,0.0491043329238891,0.002644364215672393,0.5180814401669935,0.5183384418487549,0.07068768776072235,0.5599125325679779,0.5711076855659485,0.060713358138229485,104.15079845081677,103.82135009765624,2.4599753997268077
+world_model,36,1.4482748074965044,1.4653806686401367,0.04955212498045842,0.27007457613945,0.2700395286083221,0.0011401011887412296,0.05343072522770269,0.054160539060831,0.004436308396192951,0.05138718506151978,0.0521023534238338,0.002918648761511636,0.5520685152574019,0.548346221446991,0.06711878966137087,0.592059005390514,0.6052371859550476,0.05896126351348245,97.99476554177024,97.78887176513672,2.4103264611722524
+world_model,37,1.4603810743852095,1.4705524444580078,0.03886666831617012,0.2609158158302307,0.2608284056186676,0.0010237580056190322,0.05641396275975483,0.056541621685028,0.0051271749868467146,0.052589449692856144,0.0530552007257938,0.0031888766421875716,0.5886596901850267,0.5811554789543152,0.06251745631977429,0.624869091944261,0.6394592523574829,0.056942219711722586,91.6603511463512,91.36019897460938,2.3052490151089366
+world_model,38,1.445583083412864,1.4700748920440674,0.08160809575770943,0.25177415799010877,0.2521781027317047,0.0012604347083675113,0.05921612578359512,0.0600065849721431,0.005560603924717387,0.0563673190772533,0.0564104989171028,0.0035390320123069507,0.6270109848542647,0.6159893274307251,0.05729251234197443,0.6596138910813765,0.6741161346435547,0.05434209944811939,85.36937089399858,85.0566177368164,2.2190133309235396
+world_model,39,1.4473800659179688,1.4697039127349854,0.08075579011772895,0.24198994175954294,0.241665780544281,0.0012120190482602918,0.06245997920632358,0.0630727782845497,0.006032017176564523,0.05852979184551668,0.0583397075533866,0.003915602980222885,0.667087430303747,0.6534779071807861,0.051532344651310506,0.6952203024517406,0.7095082402229309,0.051349619204571144,79.27126728404652,78.95858764648438,2.17069120262666
+world_model,40,1.4493407769636675,1.470323085784912,0.06572833988802318,0.23202403296123847,0.2320883125066757,0.0012036458383110102,0.06569067117842757,0.0654900670051574,0.006411147085674034,0.05969483032822606,0.0592914931476116,0.004219305708607066,0.7084378925236788,0.6939785480499268,0.045380733492624795,0.7307011593471874,0.7432782053947449,0.048035298935345624,73.08628845214844,72.84756469726562,2.07923501798585
+world_model,41,1.4573284062472256,1.4700312614440918,0.0370534358561458,0.22215828434987497,0.2223339676856994,0.0013724004212598267,0.07068931311368937,0.0700770542025566,0.006554823289565176,0.06379096819595852,0.0629786178469657,0.00475146840616847,0.7510665926066312,0.7382872104644775,0.039103908591640126,0.7674615762450478,0.7815163731575012,0.04401146730885997,66.92877127907492,66.76079559326172,1.904429683056744
+world_model,42,1.472347909753973,1.471879243850708,0.004512496384125374,0.21208873391151425,0.2122828066349029,0.0011106712680253046,0.07237665185874155,0.0742420554161071,0.006878926047868891,0.06591428240591825,0.0650044381618499,0.005173112327771861,0.7931421778418801,0.7826967835426331,0.03255222885666834,0.8038193420930342,0.817225456237793,0.03951663361029531,60.75834100896662,60.58831024169922,1.784008107367363
+world_model,43,1.473245078867132,1.4741942882537842,0.005057887611558761,0.2017368132417852,0.2020859122276306,0.0013095438144527392,0.07499958303841672,0.077450230717659,0.006923322805484402,0.0688668841665441,0.0676911398768425,0.005690935943467537,0.8340574665503069,0.8273907899856567,0.026183798745461617,0.8396464911374178,0.8520113825798035,0.034497810693552414,54.59273286299272,54.47153854370117,1.6672659448943383
+world_model,44,1.45997014912692,1.4646635055541992,0.02824111231732757,0.19159558550878,0.1915890276432037,0.0016183018834445798,0.07678810439326543,0.079277366399765,0.007023165993196711,0.07200149582190941,0.0705419108271598,0.006390149806645373,0.8724208149042997,0.8686785697937012,0.020170409645229478,0.8743312358856201,0.8852443695068359,0.028886887219557594,48.40534903786399,48.40412139892578,1.6125706548487995
+world_model,45,1.4271635575727983,1.4661035537719729,0.09077765048230702,0.18112877282229334,0.1811060905456543,0.0012665793972931592,0.07918936420570716,0.0816321223974227,0.006972856245226984,0.0732993219386447,0.0716978013515472,0.00687801444030462,0.9075356559319929,0.9058945775032043,0.014766805758019653,0.9061606201258573,0.9153600931167604,0.023099187008648582,42.214285070245914,42.10221099853516,1.5501295240145427
+world_model,46,1.454150069843639,1.465440273284912,0.046938897450560346,0.1707551614804701,0.1706317067146301,0.0013704254678219329,0.0800631452690471,0.0820023715496063,0.006877951209625974,0.07679086381738832,0.0749254077672958,0.007594655967727309,0.9378488930788907,0.9374494552612304,0.010142308153877509,0.9352135224775835,0.9412532448768616,0.017130158182826937,35.984771728515625,35.834083557128906,1.3928999168471727
+world_model,47,1.469654755158858,1.4709737300872805,0.007769724781248824,0.16010638258673923,0.1600614041090011,0.0012966500394825434,0.08285679058595133,0.0844769924879074,0.006863545761974524,0.08102578060193491,0.0780367851257324,0.008797688671217049,0.9631747061556036,0.9622058868408204,0.006266730232089634,0.9605456644838507,0.9635160565376282,0.011182164545195779,29.608858455311168,29.6239070892334,1.2430164643285884
+world_model,48,1.4725998748432507,1.4715147018432615,0.005562590687362959,0.14925749193538315,0.1493022590875625,0.001481298804437523,0.08466065742752765,0.086168959736824,0.0068652579897853505,0.08451796526258638,0.081049621105194,0.009985346082964086,0.9823395989157937,0.98233562707901,0.0032132131976568877,0.9805591648275201,0.98159259557724,0.005925655667429363,22.959143205122515,22.978052139282227,1.0235607160588875
+world_model,49,1.457220727747137,1.469280242919922,0.04151404224735,0.13693877377293323,0.1369135230779647,0.0014810274673513729,0.0867847089063037,0.0880544632673263,0.007269835940453091,0.08469922569665038,0.0820598602294921,0.01071361606369034,0.9947437102144415,0.9946528673171996,0.0010379004615451075,0.9936651912602511,0.9938600659370422,0.0021191341731221975,15.599080952731045,15.65239715576172,0.7342403899103459
+world_model,50,1.4238000999797473,1.4678013324737549,0.10173228326983626,0.11357312039895487,0.1135474145412445,0.0019848656695841483,0.10545582256533877,0.1074362099170684,0.009951494819751439,0.11208393221551717,0.1106475591659545,0.017324670043658678,1.0000000433488325,1.0,7.582174279590549e-08,1.0,1.0,8.429369702575069e-08,0.0,0.0,0.0

unitree_g1_pack_camera/case2/output/inference/round_summary.csv

@@ -0,0 +1,12 @@
+sample_id,scene,round_id,policy_pass_total_time_s,world_model_pass_total_time_s,round_total_time_s,latent_init_dist_to_prev_round,action_drift_vs_prev_round,psnr_full50
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,0,75.03928279876709,73.9881739616394,152.16393160820007,,,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,1,74.23106241226196,73.55208587646484,150.9755437374115,214.81951904296875,0.24937200546264648,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,2,75.07349109649658,74.42506957054138,152.62019896507263,202.9042205810547,0.24527257680892944,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,3,73.82551193237305,74.7103636264801,150.684387922287,223.25729370117188,0.0750700831413269,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,4,74.71211385726929,74.36838006973267,152.22601079940796,210.7346954345703,0.241740882396698,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,5,73.80026388168335,74.63985013961792,151.58087468147278,266.4477844238281,0.7560845464468002,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,6,74.41286873817444,74.4742693901062,151.42124199867249,239.9493865966797,0.1602541208267212,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,7,74.38386011123657,74.27649450302124,151.78351974487305,247.0240936279297,0.13166594505310059,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,8,73.25273704528809,74.35415315628052,150.77343583106995,256.5118408203125,0.06467467546463013,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,9,74.34760189056396,73.83727645874023,151.31240582466125,256.1639709472656,0.30772316455841064,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,10,73.39611744880676,74.33566379547119,150.86029887199402,246.99131774902344,0.38890862464904785,

unitree_g1_pack_camera/case2/output/inference/sample_summary.csv

@@ -0,0 +1,3 @@
+sample_id,scene,pass_type,pass_total_time_s,action_first_stable_step,state_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_state,oracle_budget_latent,latent_init_dist_to_prev_round,action_drift_vs_prev_round,round_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,psnr_full50
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,74.22499192844738,,,,44.90909090909091,46.72727272727273,46.72727272727273,47.09090909090909,,236.48041229248048,0.26207666248083117,151.4910772713748,74.22499192844738,74.26925277709961,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,74.26925277709961,,,,45.09090909090909,46.81818181818182,46.81818181818182,47.0,,236.48041229248048,0.26207666248083117,151.4910772713748,74.22499192844738,74.26925277709961,

unitree_g1_pack_camera/case2/output/sparse_8/inference/round_summary.csv

@@ -0,0 +1,12 @@
+sample_id,scene,round_id,policy_pass_total_time_s,world_model_pass_total_time_s,round_total_time_s,latent_init_dist_to_prev_round,action_drift_vs_prev_round,psnr_full50
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,0,34.508949756622314,34.254820823669434,140.46258544921875,,,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,1,34.90244817733765,34.26566195487976,141.24445700645447,216.52996826171875,0.24411267042160034,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,2,34.450573444366455,33.74520945549011,140.44232392311096,210.6576385498047,0.211828351020813,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,3,34.227173805236816,34.19744801521301,139.57855439186096,231.77476501464844,0.10076236724853516,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,4,34.42516374588013,33.859012603759766,140.9917664527893,196.30520629882812,0.24305355548858643,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,5,34.275354862213135,34.270713329315186,143.11559915542603,270.99072265625,0.9365180283784866,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,6,34.44929575920105,33.572381019592285,141.9271743297577,227.52503967285156,0.14944738149642944,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,7,34.52991771697998,34.2509183883667,141.21305179595947,256.9947509765625,0.1445719599723816,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,8,34.30908131599426,33.91361689567566,140.56905364990234,263.05572509765625,0.0568811297416687,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,9,34.12396168708801,34.13224148750305,139.65425729751587,249.03759765625,0.11548089981079102,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,10,34.32340431213379,34.3162624835968,142.47088241577148,246.99588012695312,0.5883344411849976,

unitree_g1_pack_camera/case2/output/sparse_8/inference/sample_summary.csv

@@ -0,0 +1,3 @@
+sample_id,scene,pass_type,pass_total_time_s,action_first_stable_step,state_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_state,oracle_budget_latent,latent_init_dist_to_prev_round,action_drift_vs_prev_round,round_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,psnr_full50
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,policy,34.41139314391396,,,,45.90909090909091,47.81818181818182,47.0,47.09090909090909,,236.98672943115236,0.279099078476429,141.06088235161522,34.41139314391396,34.07075331427834,
+unitree_g1_pack_camera-vid50-fs6,unitree_g1_pack_camera,world_model,34.07075331427834,,,,46.27272727272727,47.77777777777778,47.0,47.09090909090909,,236.98672943115236,0.279099078476429,141.06088235161522,34.41139314391396,34.07075331427834,

unitree_g1_pack_camera/case2/run_world_model_interaction.sh

@@ -0,0 +1,29 @@
+res_dir="unitree_g1_pack_camera/case2"
+dataset="unitree_g1_pack_camera"
+
+{
+    time CUDA_VISIBLE_DEVICES=0 "${PYTHON_BIN:-python}" scripts/evaluation/world_model_interaction.py \
+        --seed 123 \
+        --ckpt_path ckpts/unifolm_wma_dual.ckpt \
+        --config configs/inference/world_model_interaction.yaml \
+        --savedir "${res_dir}/output/sparse_8" \
+        --bs 1 --height 320 --width 512 \
+        --unconditional_guidance_scale 1.0 \
+        --ddim_steps 50 \
+        --ddim_eta 1.0 \
+        --prompt_dir "unitree_g1_pack_camera/case2/world_model_interaction_prompts" \
+        --dataset ${dataset} \
+        --video_length 16 \
+        --frame_stride 6 \
+        --n_action_steps 16 \
+        --exe_steps 16 \
+        --n_iter 11 \
+        --timestep_spacing 'uniform_trailing' \
+        --guidance_rescale 0.7 \
+        --perframe_ae \
+        --analysis_log_metrics \
+        --analysis_reference_steps 50 \
+        --head_schedule_steps 0 7 14 21 28 35 42 49 \
+        --head_skip_mode reuse_prediction \
+        --head_log_steps 40 43 46 47 48 49
+} 2>&1 | tee "${res_dir}/output.log"

unitree_g1_pack_camera/case2/world_model_interaction_prompts/unitree_g1_pack_camera.csv

@@ -0,0 +1,2 @@
+videoid,contentUrl,duration,data_dir,instruction,dynamic_confidence,dynamic_wording,dynamic_source_category,embodiment,fps
+50,x,x,unitree_g1_pack_camera,mount camera,x,x,x,G1_Dex1,30

unitree_g1_pack_camera/case3/backbone_profile_dense/inference/analysis/analysis_report.md

@@ -0,0 +1,54 @@
+# DDIM Analysis Report
+
+## Analysis Config
+
+- Stability threshold: None
+- Stability window: 3
+- Mid-step window: 15-35
+
+## Dataset Overview
+
+- Unique samples: 1
+- Step rows: 1100
+- Sample rows: 2
+- Round rows: 11
+- Scenes: unitree_g1_pack_camera
+- Pass types: policy, world_model
+
+## Five Directions
+
+### Original Early Stop
+
+- `latent_first_stable_step_policy`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_first_stable_step_world_model`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_l2_vs_full50`: mean=118.1510, median=137.4783, p90=165.7335, p95=172.5465
+
+### Direction C: Action Converges First
+
+- `action_first_stable_step`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_first_stable_step`: mean=nan, median=nan, p90=nan, p95=nan
+- `action_vs_full50_95pct_step`: mean=46.7273, median=46.7273, p90=46.7273, p95=46.7273
+- `share_action_first_stable_before_latent`: nan
+- `share_action_95pct_before_latent`: nan
+
+### Direction D: Cross-Step Similarity
+
+- `latent_delta_mid_steps`: mean=0.3528, median=0.3487, p90=0.4164, p95=0.4276
+- `action_delta_mid_steps`: mean=0.0317, median=0.0304, p90=0.0455, p95=0.0480
+- `state_delta_mid_steps`: mean=0.0317, median=0.0308, p90=0.0444, p95=0.0465
+
+### Direction A: Budget Heterogeneity
+
+- `oracle_budget_action`: mean=46.7273, median=46.7273, p90=46.7273, p95=46.7273
+- `oracle_budget_state`: mean=46.8182, median=46.8182, p90=46.8182, p95=46.8182
+- `oracle_budget_action_by_scene`:
+  - `unitree_g1_pack_camera`: mean=46.7273, median=46.7273, std=0.0000
+
+### Direction B: Round Reuse
+
+- `latent_init_dist_to_prev_round`: mean=226.9825, median=228.4588, p90=251.2165, p95=255.6648
+- `action_drift_vs_prev_round`: mean=0.2812, median=0.1662, p90=0.6238, p95=0.6602
+- `round_total_time_s`: mean=559.6007, median=588.8439, p90=625.1090, p95=625.4309
+- `policy_pass_total_time_s`: mean=231.2165, median=252.6857, p90=269.4816, p95=270.8949
+- `world_model_pass_total_time_s`: mean=239.7890, median=256.9119, p90=265.6770, p95=268.7066
+

unitree_g1_pack_camera/case3/backbone_profile_dense/inference/analysis/scene_summary.csv

@@ -0,0 +1,4 @@
+scene,pass_type,pass_total_time_s,pass_total_time_s,pass_total_time_s,action_first_stable_step,action_first_stable_step,action_first_stable_step,state_first_stable_step,state_first_stable_step,state_first_stable_step,latent_first_stable_step,latent_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_90pct_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,action_vs_full50_95pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_action,oracle_budget_action,oracle_budget_state,oracle_budget_state,oracle_budget_state,round_total_time_s,round_total_time_s,round_total_time_s,policy_pass_total_time_s,policy_pass_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,world_model_pass_total_time_s,world_model_pass_total_time_s,latent_init_dist_to_prev_round,latent_init_dist_to_prev_round,latent_init_dist_to_prev_round,action_drift_vs_prev_round,action_drift_vs_prev_round,action_drift_vs_prev_round
+,,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std
+unitree_g1_pack_camera,policy,231.21650622107765,231.21650622107765,,,,,,,,,,,44.90909090909091,44.90909090909091,,46.72727272727273,46.72727272727273,,46.72727272727273,46.72727272727273,,46.54545454545455,46.54545454545455,,559.6006835373965,559.6006835373965,,231.21650622107765,231.21650622107765,,239.78899411721665,239.78899411721665,,226.98248596191405,226.98248596191405,,0.2811613976955414,0.2811613976955414,
+unitree_g1_pack_camera,world_model,239.78899411721665,239.78899411721665,,,,,,,,,,,45.18181818181818,45.18181818181818,,46.90909090909091,46.90909090909091,,46.90909090909091,46.90909090909091,,46.81818181818182,46.81818181818182,,559.6006835373965,559.6006835373965,,231.21650622107765,231.21650622107765,,239.78899411721665,239.78899411721665,,226.98248596191405,226.98248596191405,,0.2811613976955414,0.2811613976955414,

unitree_g1_pack_camera/case3/backbone_profile_dense/inference/analysis/step_aggregate.csv

@@ -0,0 +1,102 @@
+pass_type,step,step_time_s,step_time_s,step_time_s,latent_delta,latent_delta,latent_delta,action_delta,action_delta,action_delta,state_delta,state_delta,state_delta,action_cosine_vs_full50,action_cosine_vs_full50,action_cosine_vs_full50,state_cosine_vs_full50,state_cosine_vs_full50,state_cosine_vs_full50,latent_l2_vs_full50,latent_l2_vs_full50,latent_l2_vs_full50
+,,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std
+policy,1,2.380662376230413,2.44140100479126,0.541582686726123,1.4142944487658413,1.4138203859329224,0.0021484383984422544,0.03989342092113057,0.0394162014126777,0.005337447312609944,0.03817017190158364,0.0320768542587757,0.019583009653493806,0.05996392396363342,0.052498735487461,0.060609924998752535,0.011052970520474692,0.0090840049088001,0.06042950452013075,170.6597373268821,169.39004516601562,11.4294487197575
+policy,2,4.017352884465998,4.391762018203735,1.1861475768626526,1.0213339003649624,1.0217310190200806,0.0024567828937138454,0.030183969742872463,0.0292389243841171,0.003662848656161182,0.024038863114335264,0.0234161801636219,0.0032601327954641397,0.05937811004167251,0.0544382371008396,0.060633432820552245,0.012036324648017217,0.0120233483612537,0.06094917187960682,170.7233359596946,170.29335021972656,11.105094617742372
+policy,3,4.122266856106845,4.671510457992554,1.2492590505795216,0.8495142514055426,0.8491835594177246,0.0019042913048274144,0.024150980805808797,0.0233028475195169,0.003964547134585321,0.018858069130642793,0.017807736992836,0.003598174444944763,0.062450069764798305,0.0594579800963401,0.06071553439145037,0.016193428500132116,0.0185147151350975,0.06139902092154736,169.90216341885653,168.07110595703125,11.037017541704445
+policy,4,4.412413705479015,5.128522157669067,1.3165137220462457,0.7499381845647638,0.7500395774841309,0.0020271416241545914,0.020568183030594437,0.0197155363857746,0.0039035814396075276,0.015956267629834684,0.0148289687931537,0.003215750250852455,0.06775069778615776,0.0672371089458465,0.06082244383055722,0.02261141416701403,0.0265998505055904,0.061885995271197446,169.1625615900213,167.7893829345703,10.199674945051715
+policy,5,4.495447418906472,5.103692531585693,1.3613295706748316,0.6822635748169639,0.6824440956115723,0.001260418909258719,0.016741364432329427,0.0162334702908992,0.0016007561725265203,0.014321604574268464,0.0139080937951803,0.002077126750241558,0.07432248680429021,0.0766021013259887,0.06090986389856854,0.030568414621732418,0.0359245762228965,0.06248923823081076,168.0842548717152,167.58656311035156,9.961248376480912
+policy,6,4.286584984172475,4.554376840591431,1.1852548776190672,0.6324766224080866,0.6318691968917847,0.0013462505728320158,0.0152664176103743,0.0151082212105393,0.0012230947816413758,0.0133105415681546,0.0132776191458106,0.001395284315979568,0.08171069994568822,0.0861552059650421,0.06093303013960668,0.03943565123799172,0.0461441427469253,0.0630760748117669,166.63392639160156,168.35958862304688,8.833516870372689
+policy,7,4.33748459815979,4.564946174621582,1.3863692798168286,0.5942527489228682,0.5944616198539734,0.001976783762122132,0.015312190421602893,0.014985267072916,0.001638173799170164,0.013621900823305909,0.0135518694296479,0.0014665356533676765,0.08973098343068898,0.0958151817321777,0.06098438272106316,0.04919455475597217,0.0572784952819347,0.06355077318461783,164.18314014781606,166.34677124023438,8.027096309427666
+policy,8,4.2186902436343106,4.780451536178589,1.288645832478222,0.5637131929397583,0.5641032457351685,0.0016790883821424748,0.01538657388565212,0.0147423753514885,0.0020803765192660517,0.014508365992117965,0.0140916323289275,0.0019343280256098468,0.0983921946449713,0.1056336611509323,0.06111284151095299,0.059545462824065545,0.0688173174858093,0.06396367833187915,161.15547596324575,162.53857421875,8.137841060967315
+policy,9,4.398435310883955,5.180230379104614,1.367872184846212,0.5373825810172341,0.5370771884918213,0.0015751094289736469,0.015588559464297474,0.0149363698437809,0.0019265773400665184,0.015275996263054236,0.0148359052836894,0.0021276673255401514,0.10758543488654217,0.1149793863296508,0.061177656636853754,0.07046273240650239,0.0792663246393203,0.06435803388848475,159.6385928067294,159.92527770996094,7.355049739101368
+policy,10,4.302034052935514,4.888946056365967,1.301449561747903,0.5159985748204318,0.5161250233650208,0.0012208308458228404,0.01593946022066199,0.0153064019978046,0.0016648655269842138,0.015570431439713945,0.0154294883832335,0.0018166444651003101,0.1174858405508778,0.1241016089916229,0.06124378948636174,0.08186532387679271,0.0882695466279983,0.06464688150998767,157.74214727228338,159.52822875976562,7.073231212176352
+policy,11,4.684148419987071,4.939994812011719,1.311870601643832,0.4964247400110418,0.4963581264019012,0.0012304745974528017,0.016346391883086037,0.0157323274761438,0.0015686726710732641,0.01614864246750416,0.0162970796227455,0.0012376304884104874,0.12802722914652387,0.1336388289928436,0.06122053806175511,0.09390998733314598,0.0979098603129386,0.06482495719280756,156.2006503018466,158.3176727294922,7.154127702149413
+policy,12,4.273145047101107,4.645156145095825,1.280781645333917,0.47990024902603845,0.4801351130008697,0.0008996742908775868,0.01697711375626647,0.0164263267070055,0.0020890049500293033,0.016950679942965455,0.017149893566966,0.0012110977727178523,0.13884338228539983,0.1413344293832779,0.06111893801009356,0.10661732625554904,0.108055755496025,0.06491854826714606,154.99168395996094,157.00245666503906,6.528156021503922
+policy,13,4.542992071671919,4.677780628204346,1.2000430615357585,0.4645597745071758,0.4644330441951751,0.0008753421523488961,0.01768296347423029,0.0171443223953247,0.001987090977898355,0.017695946195586107,0.0177832972258329,0.0012745236916806028,0.150162476030263,0.1499108970165252,0.06101038890091935,0.11986099031161174,0.1185173690319061,0.06496030007081756,152.89650656960228,154.0669403076172,6.78964832353932
+policy,14,4.361114566976374,4.165746450424194,1.4476579026619352,0.4507998688654466,0.4507092833518982,0.001148390947731996,0.018548831682313527,0.0180491916835308,0.0018748463598881049,0.01874042657965959,0.0187033209949731,0.0013111508043958716,0.16187893463806668,0.1586249023675918,0.06102617630306333,0.1336405202746391,0.1349441856145858,0.06499711722733852,151.06254716352984,151.5512237548828,6.397771345287218
+policy,15,4.34663013978438,4.718540668487549,1.3759562390189437,0.437546814029867,0.4371760189533233,0.0013334699989280738,0.019253784113309545,0.018957681953907,0.001517591044523947,0.019470439038493364,0.0194622557610273,0.0013085747503791035,0.17409252443096854,0.1679667830467224,0.06099290289328309,0.1478756215762008,0.1518889218568802,0.06506348480913784,149.530517578125,148.61514282226562,6.578709363317128
+policy,16,4.851059805263173,5.363605499267578,1.546863468789342,0.4263095936991952,0.4265479445457458,0.0007359383716435437,0.019992871209979,0.0196767337620258,0.0013134368477212236,0.020573206923224653,0.0207148194313049,0.0012400297638149643,0.18697640977122562,0.1784685254096985,0.06100492366568693,0.16268579661846158,0.169473722577095,0.06507147085861935,147.874980579723,148.01016235351562,6.066549560034442
+policy,17,4.611436583779075,4.841576099395752,1.5418576205612777,0.415495211427862,0.4153204560279846,0.0010887353063793632,0.02091397328132928,0.0205698162317276,0.0012625873576255988,0.02161299335685638,0.0217533241957426,0.0012046574690941386,0.20022738386284217,0.1919917166233062,0.06117531480525082,0.1780587170611728,0.187808945775032,0.06503634882526531,146.553242076527,145.9924774169922,5.604666616949082
+policy,18,4.650238860737193,4.870306015014648,1.246547673549721,0.40552167188037525,0.4056895673274994,0.0008204037957580346,0.021648015149614983,0.0213276743888854,0.001105919888648144,0.022350997240705873,0.022369222715497,0.0012364848723696578,0.21419291469183832,0.2076755315065384,0.06136309946111001,0.19381767037239941,0.2064991593360901,0.06497108980450088,144.86958035555753,144.37498474121094,4.9296349802721116
+policy,19,4.245500629598444,4.738556623458862,1.4302690861917486,0.39656902985139325,0.3964414298534393,0.0007120039526937954,0.02304079028015782,0.0227679833769798,0.001277709554205896,0.023567383431575453,0.0235518626868724,0.0012767681038816216,0.22836149416186588,0.223897099494934,0.061481838091152805,0.21018501574342896,0.2259309291839599,0.06488431766960985,143.78720647638494,143.2709503173828,4.586066345176296
+policy,20,4.444193341515281,5.092517137527466,1.3926681132622294,0.3868722753091292,0.3866529464721679,0.0009725648801149819,0.023886837573213946,0.0235368832945823,0.0011709715856123278,0.024389537220651406,0.0243271104991436,0.001341701289877731,0.24327235465699973,0.2406058758497238,0.061617539000575365,0.2270290580662814,0.2460137903690338,0.0648149842712134,142.74720486727628,141.80227661132812,4.577040859157341
+policy,21,4.386648980053988,5.009429693222046,1.363857340237241,0.37908863479440863,0.3790760338306427,0.0010430826405804427,0.024882349439642612,0.0245757140219211,0.0011766313773369761,0.02588456293398679,0.0258131921291351,0.0013265805314550527,0.2588122649626298,0.2577861249446869,0.06179225291083729,0.24465590173547913,0.2668616771697998,0.06471350337039469,141.845703125,141.22067260742188,4.3996323181232615
+policy,22,4.319781476801092,5.048331022262573,1.3802578345556755,0.37081768025051454,0.3709299266338348,0.0007739942996079611,0.026193853806365554,0.0259468480944633,0.001213840728079604,0.027134262025356237,0.0271316897124052,0.0013551035476557995,0.2749268060380762,0.2756619751453399,0.0619689739555047,0.26295746862888336,0.2885586619377136,0.06458037662957057,141.11494168368253,140.4208526611328,4.335681809792902
+policy,23,4.593331185254184,5.236274242401123,1.3275380820338754,0.36260494047945196,0.3623712360858917,0.0008190675532954606,0.027484248984943702,0.0273963455110788,0.0012314467330793088,0.0283150161531838,0.0283326767385005,0.0013392103433690942,0.29147974469444965,0.2930734753608703,0.06216866563161143,0.2819277468052777,0.3109440803527832,0.06441482173019429,140.2742392800071,139.86187744140625,3.8695717251843704
+policy,24,5.0160212950273,5.633982181549072,1.492315421534178,0.35600360686128785,0.356194794178009,0.0006575709755798822,0.0287465236403725,0.0285717919468879,0.0011884048444063399,0.029680302027951546,0.0295587237924337,0.001357358438982101,0.30927984551949933,0.3099324107170105,0.062333690068765955,0.3016752830960534,0.3331969976425171,0.06424982576074391,139.4898126775568,138.93145751953125,3.637206712630539
+policy,25,4.802242950959639,5.09297776222229,1.3332438475330286,0.34881880337541754,0.3490033745765686,0.0004511000411320036,0.030162663283673164,0.0299511235207319,0.0013024558069855982,0.031122536313804662,0.0308958888053894,0.0013711095591744343,0.3281232199885628,0.3276771306991577,0.062287652184931354,0.32224926758896216,0.3552346229553222,0.06401714833319155,138.64874545010653,138.25657653808594,3.3541121994740877
+policy,26,4.813915404406461,5.1604905128479,1.1356238785527322,0.3419828631661155,0.3421078026294708,0.0006603708735225903,0.031617173925042104,0.0314201898872852,0.001375520649966228,0.03240807676179838,0.0323491431772708,0.0013898252291039697,0.3475651768120852,0.3457567393779754,0.06225008594074336,0.3436255766586824,0.3781275153160095,0.06375825453650781,138.031757701527,137.58840942382812,3.224431057876386
+policy,27,4.368636651472612,5.008497953414917,1.4205146555464678,0.3357629613442854,0.3356943130493164,0.0005567016840524652,0.033215382559732914,0.0330285131931304,0.0014825439021736776,0.033993283456022046,0.0340738892555236,0.0014185863282812853,0.3679606941613284,0.3648787438869476,0.062030980628245386,0.36594438552856445,0.4017867147922516,0.06342569804765069,137.51301297274503,137.48892211914062,3.0607299682352593
+policy,28,4.534430287101052,5.133967638015747,1.2592549284773933,0.3289194350892847,0.3289177119731903,0.0009514299831212895,0.03470892154357646,0.0344643704593181,0.0016119507708267208,0.0355001884427937,0.0356693007051944,0.0015011545571736915,0.389326583255421,0.3848195374011993,0.06166297511632919,0.3892082165588032,0.4227328896522522,0.0629848693283037,136.92720725319603,136.971923828125,2.862472807497929
+policy,29,4.757577462629839,5.214699029922485,1.4943077191607803,0.3225756910714236,0.3228643238544464,0.0007955021505878752,0.036326250569386875,0.0359343774616718,0.001641171756995819,0.037107751112092566,0.0370557308197021,0.0015285782395413731,0.4120210382071408,0.4059362411499023,0.06098504919378585,0.41347925771366467,0.4414198100566864,0.06246565938610993,136.25204745205966,136.38461303710938,2.6919436675587547
+policy,30,4.811137936332009,5.319025993347168,1.2077632932875766,0.31666511297225947,0.3164407610893249,0.0008908328901844494,0.03817518834363325,0.0378871411085128,0.0018152697028262664,0.039139174263585645,0.0388884581625461,0.001619654966819521,0.4356038109822707,0.4277892112731933,0.059973987597738226,0.4389731938188726,0.4612630009651184,0.06177581463242909,135.7519170587713,135.8135986328125,2.5421817759608922
+policy,31,4.821598464792425,4.899641036987305,1.174137722004987,0.31148783727125684,0.31135293841362,0.0008010109738308051,0.03998917883092703,0.0396871455013752,0.001846099564326508,0.040883624418215275,0.0404579415917396,0.0017520350856787925,0.4605525921691548,0.4511896073818207,0.059013804263764374,0.46558957208286633,0.4820389151573181,0.06095196863453568,129.45022929798472,129.62844848632812,2.4053850154855088
+policy,32,4.955651326612993,5.39706826210022,1.2825437333065932,0.3037239990451119,0.3038488030433655,0.000680703131185027,0.042273248122496984,0.0420779772102832,0.0022244315296006717,0.04210943525487721,0.0415312461555004,0.00184747315730652,0.4862545918334614,0.4754481911659241,0.05786149206172697,0.49304402416402643,0.5036220550537109,0.06006459167840216,123.248159235174,123.44611358642578,2.3607473037271953
+policy,33,4.807590267874978,5.503448247909546,1.3443983246439242,0.29539818926291034,0.2954290807247162,0.0009959418129323152,0.044100029902024655,0.0439146384596824,0.0020702669828953917,0.04420756447044283,0.0435491278767585,0.002017416341566527,0.5141065716743469,0.5016157627105713,0.0563753291378469,0.5217923305251382,0.5264453887939453,0.05899675911506213,117.10901295055042,117.23194885253906,2.2315259467821296
+policy,34,4.7405454895713115,5.256289958953857,1.2211906211637598,0.2871642383662137,0.2869498133659363,0.0006625785506041712,0.04620384932918979,0.04602075740695,0.0022970677126611916,0.04621206630359991,0.0454258099198341,0.0021619731984360035,0.5429608009078286,0.5289064645767212,0.05478619910668787,0.5517389503392306,0.5505180358886719,0.057724558223585146,110.83411754261364,110.73796844482422,2.200723821196813
+policy,35,4.8246071988886055,5.152066946029663,1.2825231190302535,0.27888934991576453,0.2787675559520721,0.0012822400997129888,0.048534341156482655,0.0482332669198513,0.0025636847857765177,0.04813042994249947,0.0470250733196735,0.002351723157774234,0.5730243271047418,0.5573737621307373,0.0530125314964134,0.5827630975029685,0.5757917165756226,0.05619385664163848,104.62270008433948,104.51995086669922,2.027768465321225
+policy,36,4.955069346861406,5.519265413284302,1.4112487180624358,0.2696909362619573,0.2693997919559479,0.0009001476887071999,0.05106754025275054,0.0508812889456748,0.0027335473449138187,0.05120867219838225,0.049693826586008,0.0027298500319434603,0.6044634580612183,0.5874694585800171,0.05098130040112226,0.6154280250722711,0.6068510413169861,0.05411261369600538,98.42063002152877,98.2171401977539,1.9906164929381902
+policy,37,4.7650146484375,5.186834096908569,1.1140044477965456,0.2604475834152915,0.2602835297584533,0.0008038996565517731,0.05369132994250813,0.0540202967822551,0.002894223687384794,0.05217225036837835,0.0510293580591678,0.0028366312470402633,0.6378475048325278,0.6202614903450012,0.04843071763634275,0.6483965136788108,0.6398486495018005,0.05194550207698849,92.09824579412287,91.82431030273438,1.9734453453669998
+policy,38,4.874200235713612,5.024654150009155,1.1381593685369071,0.25149029493331904,0.2513195872306824,0.0009999253181738436,0.05606721443208777,0.0564273372292518,0.003150326134148752,0.05569560893557284,0.0549632795155048,0.0033310480765264645,0.6723501844839617,0.6554672718048096,0.04544466049797355,0.6829756715080955,0.67472904920578,0.04906886327274817,85.82743141867898,85.60066223144531,1.949808032265999
+policy,39,5.009584708647295,5.174219369888306,1.1811262947862227,0.2418465966528112,0.2418252974748611,0.0011894211721312255,0.05897240035913201,0.0599682740867137,0.0035627918259868614,0.057560757479884365,0.056814718991518,0.0036230940232781232,0.7081743532961066,0.693757176399231,0.04185765902640528,0.7179932756857439,0.7099788188934326,0.045863835499359754,79.74040499600497,79.55411529541016,1.8624035978132791
+policy,40,4.877714525569569,5.219934463500977,1.100921844524881,0.23208030922846357,0.2319167107343673,0.0010777016060087526,0.06177945435047145,0.0631598606705665,0.003970350155229003,0.05858897519382558,0.0576555244624614,0.0038750994858483614,0.7447838945822283,0.7327991724014282,0.037776759442364506,0.7526359937407754,0.7446768283843994,0.04242795424530628,73.57139726118608,73.59123229980469,1.8349843884243335
+policy,41,4.727325331081044,5.006701946258545,0.9767662652194806,0.22194951908154917,0.2221188247203827,0.0011349956232010491,0.06619690697301514,0.0673195272684097,0.004520878959155858,0.06222958524118765,0.0614445284008979,0.004446536758516849,0.7820957357233221,0.7713624238967896,0.03331132059053609,0.7880134853449735,0.780337929725647,0.038255944571623984,67.40430866588245,67.43851470947266,1.8184118098596465
+policy,42,4.908030076460405,5.077185153961182,1.0845275540646921,0.2116578952832655,0.2118271738290786,0.0009935721008423222,0.06785158677534618,0.0696453228592872,0.005022372922747044,0.06408537958155974,0.0632410869002342,0.004852935947155101,0.8190687840635126,0.810761570930481,0.02837859787820384,0.8226447213779796,0.8165727853775024,0.033784786994660934,61.1750207380815,61.23821258544922,1.812018700250013
+policy,43,5.032526037909768,5.348829984664917,1.1547953853614703,0.2016072327440435,0.2017250955104828,0.0009877936454333511,0.0702517838640646,0.0720702409744262,0.0055376944314737145,0.06653266433965072,0.0660029575228691,0.005435586443460142,0.8550188812342557,0.8492221832275391,0.023237661327779543,0.8562712994488803,0.8535821437835693,0.028920346573870794,55.039310455322266,55.25835418701172,1.8328126169931669
+policy,44,5.191218549554998,5.621124267578125,1.1814409204807697,0.19093608449805863,0.1909235715866089,0.0010272308029926958,0.07197818126190789,0.0738246589899063,0.005937746732717742,0.0691897594793276,0.0687492340803146,0.006006961402072175,0.8887581066651777,0.8849911689758301,0.018218433451395198,0.8883262385021556,0.888175904750824,0.023745739886244205,48.825135317715734,49.09001159667969,1.8135309985521677
+policy,45,5.060894250869751,5.413084506988525,1.1282336199333933,0.18100758845155884,0.1809582114219665,0.000994496786733235,0.0741657221859151,0.0763023570179939,0.006249535501236141,0.07021897353909229,0.0698379352688789,0.006413494683002127,0.9195944450118325,0.9165513515472412,0.013476723387178011,0.9173889756202698,0.918834149837494,0.018640412691943952,42.60464859008789,42.85799407958984,1.7372615030798464
+policy,46,4.9386608383872295,5.266998529434204,1.1007331048217883,0.17041375420310279,0.1700872927904129,0.0009691128268017833,0.07481189004399556,0.0770504847168922,0.006435596031731268,0.07305758988315404,0.0734641626477241,0.0071789160712951335,0.9461199370297518,0.9451377391815186,0.009283734379430057,0.9434904727068815,0.945602536201477,0.013536202933917541,36.28016489202326,36.45746612548828,1.584704067163711
+policy,47,5.085831121964888,5.24397873878479,1.1576885103106176,0.15976025164127347,0.1599758714437484,0.0010365037037389582,0.07713298093188889,0.079283133149147,0.0065933765163542915,0.07653772424567824,0.0777846649289131,0.00818926604621114,0.968178781596097,0.96835196018219,0.005699818499982981,0.9658823013305664,0.9679090976715088,0.008673177641887535,29.784864252263848,29.960580825805664,1.4092411426303904
+policy,48,4.80556828325445,5.23617434501648,1.1935121350035067,0.14887292276729233,0.1490814387798309,0.0010870975155640246,0.07835294170813122,0.0803812071681022,0.006787764937893993,0.07915468243035399,0.0796021074056625,0.009175547367602887,0.9847689650275491,0.9852114915847778,0.0028689876643410027,0.9832953810691833,0.9847090244293212,0.004521617006090728,23.075346513227984,23.132247924804688,1.1988878437284138
+policy,49,4.8093796860088,5.022816181182861,1.1513628995112177,0.13655048066919495,0.1363487988710403,0.0013178552769897666,0.07985884222117333,0.0809646174311637,0.007252202439408063,0.07915049723603504,0.0787261649966239,0.009794229849653861,0.9954712336713617,0.9956791400909424,0.0009023152689976716,0.9946055195548318,0.9952359795570374,0.001580523210549921,15.612876371903853,15.678373336791992,0.8141453611561115
+policy,50,4.954349777915261,5.356104135513306,1.203074507735412,0.11318790506232865,0.1128654554486274,0.0018904817665993832,0.09636235440319231,0.0955251827836036,0.009557327116620527,0.10407706011425359,0.1008240580558776,0.014747060602726911,0.9999999349767511,0.9999999403953552,9.777540358224394e-08,0.9999999945813959,1.0,7.750687618660017e-08,0.0,0.0,0.0
+world_model,1,2.4772726622494785,2.6260852813720703,0.4633811981794726,1.4124032692475752,1.412093997001648,0.002457663891915528,0.04069613157348195,0.0417177006602287,0.00375100830758644,0.031324904750693876,0.0296769831329584,0.004123942941606412,0.030547814274376098,0.0170293115079402,0.07325087245379798,0.03789447662843895,0.0133213605731725,0.07681115888595182,168.9413493763317,167.04258728027344,8.870859674138815
+world_model,2,4.588764840906316,4.935030698776245,1.1437340301634655,1.022136471488259,1.0221750736236572,0.0019465726925450286,0.029069267044013147,0.0288726761937141,0.0026851660834261897,0.025301381268284493,0.0228177830576896,0.006132891700030667,0.030541637404398473,0.0147097278386354,0.07415092426376077,0.03911464381963011,0.0134562943130731,0.07769090991148832,168.9006805419922,166.25335693359375,8.813890422183835
+world_model,3,4.814267006787387,5.164716720581055,1.080139639616605,0.8511152809316461,0.8514657020568848,0.0020225865131282376,0.021794746545228057,0.0215661283582448,0.002482135518973717,0.019999342885884317,0.0183340515941381,0.004875596783062964,0.033488688651811,0.0179222803562879,0.07526511375664556,0.04335922510786488,0.0175474118441343,0.07784427391511073,168.27509654651988,167.83848571777344,8.496368107474087
+world_model,4,4.837730776179921,5.266719579696655,1.1864552167442637,0.7505818063562567,0.7506167888641357,0.0017866412140438678,0.018212452700192228,0.0180761832743883,0.001783864170658293,0.01678404423662203,0.0164924133569002,0.002028387959830416,0.03834511407397009,0.0240975469350814,0.07616942533630729,0.04979974776506422,0.0252106338739395,0.0775102112272449,167.7500887784091,167.43145751953125,8.250289902982816
+world_model,5,4.933083707636053,4.993900060653687,1.1563012493193514,0.6825656890869141,0.682481586933136,0.0016335568798638764,0.0163001893088221,0.0162542387843132,0.0014766742245574285,0.015678072686899747,0.0150940055027604,0.0023224130829244925,0.0445173049176281,0.0316356867551803,0.0769019204121999,0.057512312026863716,0.0350181758403778,0.07694999404317358,166.15844171697444,166.4280242919922,8.916397026186825
+world_model,6,4.6008995663036,4.707776308059692,1.1297094701275836,0.6334039460528981,0.6324781179428101,0.002084275834032274,0.015360233289274262,0.0155090689659118,0.001334267749440454,0.0143509342081167,0.0138716101646423,0.0023948690410499794,0.05152712863954631,0.0399010330438613,0.0774343300964602,0.06599576771259305,0.0457124225795269,0.07626917998525844,164.33296342329547,166.19485473632812,8.341476358254052
+world_model,7,5.032595699483698,5.253226518630981,1.1371081874540119,0.5949551625685259,0.5946858525276184,0.0014701912777972256,0.015421443872831037,0.0152621762827038,0.0014150637585943749,0.014295721257274735,0.0133011536672711,0.0025565200064464485,0.0589852597225796,0.0486289113759994,0.07800423922193517,0.07520853897387327,0.0570916235446929,0.07549901931392759,162.1347226229581,162.78929138183594,8.436401152756197
+world_model,8,5.090328910134056,5.4763147830963135,1.210936623228668,0.5634629401293668,0.5636498332023621,0.0014825157663435667,0.015587496825239827,0.0154073508456349,0.0014829289325611116,0.014804785901849882,0.014084218069911,0.002187404825633583,0.06707560287957841,0.0576387904584407,0.07855575463840891,0.08502513593570747,0.0690293461084365,0.07463214613858797,160.50648498535156,160.97909545898438,8.741872924310313
+world_model,9,5.037052739750255,5.324824571609497,1.1248074097885892,0.5372636101462624,0.537748396396637,0.0011631932501631852,0.015483319082043338,0.0150259742513298,0.0014763234848261821,0.015057365071367084,0.0147182513028383,0.0015073305287420077,0.07574839475141328,0.0672002136707305,0.07902963580955878,0.09545684775168244,0.0814889371395111,0.07373243647889993,158.46505598588422,157.92214965820312,8.22491356028657
+world_model,10,4.6527089205655185,4.874844074249268,1.203397904154107,0.5154841433871876,0.5150421261787415,0.0012415871970163442,0.01559340953826899,0.0152573743835091,0.0014127163758630731,0.0151289512657306,0.0150147480890154,0.0018122397128643656,0.08522028442133554,0.0774843543767929,0.07943486863408895,0.10654258677227926,0.0943956971168518,0.07284864729965178,156.2073516845703,156.52951049804688,7.366497223582135
+world_model,11,5.050711675123735,5.32934045791626,1.1142941572340692,0.49662127819928253,0.4963216781616211,0.0012074165231440778,0.0159081922505389,0.0153419086709618,0.00135108146682072,0.015862169844860333,0.0160740502178668,0.0013647432992026996,0.09527192213995889,0.0881689488887786,0.07978880333793494,0.11835134537382558,0.1078644543886184,0.07214613331265148,154.1974764737216,154.14584350585938,6.802427454241447
+world_model,12,5.0107127969915215,5.480522871017456,1.1911792164117039,0.4795596193183552,0.4793055653572082,0.0013786704369920522,0.016203014857389654,0.0158650558441877,0.0013214168189378426,0.01663460506295611,0.0166973024606704,0.0013492486214416047,0.10588396255942904,0.0990850180387497,0.08022048983078035,0.13087093186649404,0.1217457428574562,0.07151675554585432,152.3022225119851,152.15127563476562,6.132136723108185
+world_model,13,5.121617663990367,5.587911128997803,1.1588778707119036,0.4639504741538655,0.4640955328941345,0.0010126082662225784,0.016975870186632273,0.0165792852640151,0.0013585680898495578,0.0173192407766526,0.0173036605119705,0.0014320493770926806,0.1169683555649085,0.1104472950100898,0.08082887464580593,0.1438616314394907,0.1359855234622955,0.07085395524109815,150.74200439453125,149.87294006347656,6.057937901769739
+world_model,14,4.920238083059138,5.18118953704834,1.222186028696898,0.45046493681994354,0.4501868784427643,0.0013117565086260697,0.0180127969519658,0.0177614279091358,0.0014429405306807363,0.018324904821135718,0.0184523686766624,0.001485379376591726,0.12842326340350232,0.121901050209999,0.0815451242657217,0.1574339414523406,0.1507323086261749,0.07016672039432788,149.08225042169744,147.71583557128906,5.942993849851055
+world_model,15,4.9171084057201035,4.98837423324585,1.218442546664169,0.4382816634394906,0.4385071396827698,0.0009232861629472268,0.018824361434037,0.0185197480022907,0.0012385880106404196,0.01901468888602469,0.019109645858407,0.0014832487086463976,0.1405601281334053,0.1337485760450363,0.08204180289071585,0.17137388715689827,0.1657336205244064,0.06951096162254558,147.807504827326,146.6032257080078,6.1300975607326205
+world_model,16,4.939292734319514,5.021778345108032,1.14165544182925,0.4266241030259566,0.4266462028026581,0.0010287450573323613,0.019778944213281916,0.0194871183484792,0.0012541059782059973,0.020121345465833444,0.0202668253332376,0.0014965586771063747,0.15327224457128477,0.1461504697799682,0.08236093603202554,0.1858679889277978,0.1813174337148666,0.06881955687597428,146.4139043634588,145.67523193359375,5.961963773465782
+world_model,17,4.962915702299639,5.230148792266846,1.103332875925909,0.41554778543385584,0.4155460894107818,0.0006550268582073158,0.020688755776394466,0.020557090640068,0.0012749980794489607,0.02116495607928792,0.0211744960397481,0.001563105795299829,0.1666191812943328,0.1599164605140686,0.08259432062206087,0.20088511840863657,0.1957357227802276,0.0681005123769131,145.5342351740057,145.08901977539062,5.79038353965572
+world_model,18,4.892659382386641,5.2655930519104,1.187008359547742,0.4060117385604165,0.4060111343860626,0.0003950187417536812,0.02144301276315337,0.0215845629572868,0.0012847020327515191,0.021855192936279497,0.0218735728412866,0.0015090120661513118,0.1807078509168191,0.1745617836713791,0.08270701176574037,0.21622793105515561,0.2094273716211319,0.0674174988395342,144.09886863014916,144.51260375976562,5.348353112718343
+world_model,19,4.723549409346147,4.97363018989563,1.3150893446755383,0.39582287723367865,0.395850658416748,0.0008982450620106717,0.022741166536103544,0.0224764291197061,0.001204178363978115,0.023033933714032118,0.0230550020933151,0.0015407086278670186,0.19484091787175695,0.1894541382789611,0.0829718936467968,0.23217284340750083,0.2259703874588012,0.06676109659281883,143.02939813787287,143.82711791992188,5.198856994175708
+world_model,20,5.034282099116933,5.2240519523620605,1.1525268411209983,0.3870746059851213,0.3869898319244385,0.00045280298162217594,0.023740503259680447,0.0236174017190933,0.0014024631780584255,0.023808748884634492,0.0238881148397922,0.0015444328298981852,0.20978705381805243,0.2045099735260009,0.08293436792353304,0.24852788990194144,0.243560642004013,0.06610120788822259,142.30528259277344,142.91085815429688,5.155826396009903
+world_model,21,5.135964306918058,5.472920656204224,1.136127446329099,0.37906705791299994,0.3787281513214111,0.000839016301694403,0.024921110577203972,0.0247568525373935,0.0016157658619052233,0.02526675249365239,0.0253008734434843,0.001581960175408436,0.22530335594307294,0.2196742296218872,0.08291130003278482,0.26561422510580585,0.2619453966617584,0.06541789990761261,141.47780123623934,141.8600311279297,5.07521546938977
+world_model,22,4.9428016705946485,5.354492902755737,1.1880017753040124,0.37078800255602057,0.370766818523407,0.0007952269942321572,0.026379755952141447,0.0260683856904506,0.001608699397474075,0.026498705318028226,0.0265805963426828,0.0015889499278207253,0.24132458459247239,0.2353464663028717,0.08278060910096924,0.2833121811801737,0.2807280719280243,0.06469473103505267,140.6860448663885,140.71827697753906,4.908735311316538
+world_model,23,4.979956366799095,5.276267051696777,1.1844509064966688,0.36288704655387183,0.3627218306064605,0.0009210684202446669,0.027785677632147576,0.0273771472275257,0.0017094181332869616,0.02766189334744752,0.027649188414216,0.0016235942930448684,0.2578049627217379,0.2514077425003052,0.08288921968578288,0.30163155360655347,0.3000853657722473,0.06390466456204294,139.83792252974078,139.9789276123047,4.602857860632147
+world_model,24,5.071351874958385,5.461961269378662,1.2186496996922964,0.35558476773175324,0.3553908467292785,0.0008089210319027462,0.0289607630534605,0.0289991367608308,0.0017401378516227097,0.028945115479555947,0.0289651826024055,0.0017105791036972579,0.2755074636502699,0.2689995169639587,0.08267489373037475,0.32066073878244916,0.3202436566352844,0.06304429517504696,138.9628351384943,139.2001953125,4.38988801424008
+world_model,25,4.815841197967529,5.36384129524231,1.1803513068896276,0.3483759408647364,0.3485066592693329,0.0004712920337327591,0.03038015860048203,0.0305563770234584,0.001813130000992769,0.03033043020828199,0.0303567294031381,0.001753163978961418,0.294265634634278,0.2879073917865753,0.08236780614028984,0.3404816009781577,0.3412869572639465,0.06217648227010578,138.23024541681463,138.46278381347656,4.020007337395506
+world_model,26,4.914045117118142,5.384947538375855,1.2953615314855789,0.341809405521913,0.3418368101119995,0.001040485883566752,0.03190619667822661,0.0319111868739128,0.0018373233874340912,0.0315490691837939,0.0316865593194961,0.0018062918340623847,0.3136223310774022,0.3075994849205017,0.08221572378161891,0.3609896275130185,0.3630464076995849,0.06120919487150544,137.63728748668325,137.95166015625,3.7660400014420854
+world_model,27,5.035937612706965,5.112766265869141,1.1776955482888005,0.3348960036581212,0.3350080847740173,0.0010042699521963787,0.03358456509357143,0.0338612310588359,0.0018424645990354372,0.03311696902594778,0.033239722251892,0.00185993216681556,0.3340018201958049,0.3287181556224823,0.08207020455081274,0.38235821507193823,0.3857570886611938,0.060164366329054295,137.04050237482244,137.14755249023438,3.6928644816480074
+world_model,28,5.028671459718184,5.133925676345825,1.2780196537896702,0.3288316401568326,0.3287160396575928,0.0008137902281180491,0.03513622859662224,0.0352833829820156,0.0018635662554352319,0.034590359607880726,0.0345665700733661,0.0019439252571908927,0.3554344908757643,0.3516155779361725,0.08188443873935997,0.40458517453887244,0.4087463617324829,0.059039967961300815,136.5198322642933,136.48324584960938,3.536415509648704
+world_model,29,4.91842380436984,5.1893510818481445,1.2134013344677785,0.32263135639103974,0.3227916359901428,0.0009935641271685452,0.036820090968500446,0.0372705310583114,0.0019157351612340244,0.036167550832033116,0.0361042059957981,0.002025591504346536,0.37831414423205634,0.3759738206863403,0.08141903265836435,0.42779106172648346,0.4324239790439605,0.057834882034773265,135.82081742720172,135.85202026367188,3.467110970663888
+world_model,30,5.106633403084495,5.491691827774048,1.1821930558051839,0.31670329787514423,0.3169687986373901,0.0009193396285672977,0.03874675557017321,0.0392829775810241,0.0018682288904286538,0.03820227391340511,0.0380441211163997,0.0020682663124523444,0.40214644236998126,0.4016953706741333,0.08103820997950145,0.45220245014537463,0.457308679819107,0.056545307334280336,135.26103626598012,135.26927185058594,3.2359504699223396
+world_model,31,4.811611674048684,5.094662427902222,1.2265523022586147,0.31152299588376825,0.3114700615406036,0.0008357155000593058,0.04072053391825064,0.0415845997631549,0.0020991201837283263,0.03996928510340772,0.0398033000528812,0.00216472348852389,0.4276374307545749,0.4287772178649902,0.08022330304121951,0.47771289673718537,0.4832184910774231,0.05513682245853667,128.9315095381303,128.96115112304688,3.2572223658946577
+world_model,32,4.762716575102373,5.015646457672119,1.0769884621946977,0.30359881845387543,0.3040041029453277,0.0010768252713368152,0.043077714402567205,0.0431665927171707,0.0021949497587082857,0.04126793112267142,0.04104470834136,0.0021892222104031474,0.45416016199372033,0.4549955129623413,0.07940588728119641,0.5040509294379841,0.5104079246520996,0.053690133727726616,122.74886460737748,122.66412353515624,3.1729853752407076
+world_model,33,4.4435717409307305,5.0545947551727295,1.3338526071962873,0.2955843399871479,0.2955531477928161,0.0009975414791941542,0.045203314586119175,0.0459014847874641,0.002848557929123764,0.04335263710130341,0.0428242497146129,0.002285311569467592,0.48304073918949475,0.4828449189662933,0.07748654071144045,0.531664948571812,0.5390397310256958,0.05212307182097492,116.5165356722745,116.45047760009766,3.0580948563853863
+world_model,34,4.702559015967629,5.232410907745361,1.2673377620690809,0.28740227493372833,0.2875078618526459,0.0008757606989314342,0.047460238364609775,0.0483389236032962,0.002860846428998898,0.045352392237294745,0.0448821485042572,0.002400065646042707,0.5132270048965107,0.5150481462478638,0.07546497986832192,0.560425644571131,0.5687294006347656,0.05037034088334785,110.2532168301669,110.08950805664062,3.031555569151996
+world_model,35,4.910938544706865,5.273892164230347,1.3460566282554576,0.278427086093209,0.2782846093177795,0.0007285301242817567,0.04998143796216354,0.0507170930504798,0.0031207586904566918,0.04730376431887797,0.0467439629137516,0.0025374985520163463,0.5449797240170565,0.5481972694396973,0.0730576581284355,0.5902385765855963,0.5992961525917053,0.04839746794681975,104.06160319935192,103.84439086914062,2.9887733642368675
+world_model,36,4.783166126771406,5.401458501815796,1.2524155132435961,0.27031036127697333,0.270441472530365,0.0007438008267988205,0.05259096249938006,0.0531465522944927,0.0034521181013809185,0.05041778290813615,0.049702201038599,0.0027646660000093955,0.5781964415853674,0.5825752019882202,0.070176573190219,0.6217330152338202,0.6315561532974243,0.04613660347805826,97.855825250799,97.38742065429688,2.897257229306152
+world_model,37,4.648333831266924,4.868606805801392,1.1167099499964384,0.2610315219922499,0.2613421380519867,0.001162534274888696,0.05544891750270666,0.055097758769989,0.004342124592322422,0.051443698054010174,0.0505328029394149,0.002842930712398791,0.6133698387579485,0.61818927526474,0.06623606872776741,0.6535887338898398,0.6638416647911072,0.04381833133245995,91.62903733686967,91.27247619628906,2.7797085168391438
+world_model,38,4.811313065615567,5.214028358459473,1.1877347707091497,0.2514737085862593,0.2516121864318847,0.0011393127100485566,0.058026051995429075,0.0572158023715019,0.004787305185333226,0.054949085820804905,0.0541835501790046,0.003164539186418428,0.6497948332266374,0.6550724506378174,0.06173128459177867,0.6870355660265143,0.6976069211959839,0.04100373462208296,85.35032931241122,84.9329605102539,2.704297101404805
+world_model,39,4.755873333324086,5.303229808807373,1.3015866943683905,0.2417475269599394,0.2418097257614135,0.001085692574350814,0.06082476878708054,0.059891253709793,0.005241590288301536,0.056810456920753774,0.0561013296246528,0.003367155789979665,0.6875223517417908,0.6931766271591187,0.056661086393873704,0.7209509015083313,0.731539785861969,0.03801295848853494,79.3028134432706,78.83509063720703,2.5676890611057432
+world_model,40,4.705951235511086,5.259236574172974,1.489435251731432,0.23207700794393366,0.2323889434337616,0.001143536174678739,0.0636763444001024,0.0626358389854431,0.005689209573549386,0.057808818803592114,0.056958507746458,0.0035600810727225733,0.7261427911845121,0.7318187952041626,0.05095700052681276,0.7545402483506636,0.7646765112876892,0.03492720863993108,73.15307686545633,72.64688110351562,2.509586329263128
+world_model,41,4.820034070448442,5.327846050262451,1.551689167113277,0.2219962816346775,0.2217660993337631,0.001177181096158895,0.06845207613977514,0.068399854004383,0.006113041797920698,0.06150785054672845,0.0602356269955635,0.004011159842403013,0.7658631313930858,0.7699307203292847,0.04459192107683944,0.7889649976383556,0.7981904745101929,0.03131238161299526,67.03927924416281,66.38236999511719,2.422610066235947
+world_model,42,4.557164083827626,4.842182636260986,1.4215606238357223,0.21196154301816764,0.2120925188064575,0.001219411519846473,0.07018301161852745,0.0699777007102966,0.006765810782613124,0.06337224793705068,0.0618951916694641,0.004293874897947139,0.8053750991821289,0.8089163899421692,0.03756453120551996,0.8227699128064242,0.828944981098175,0.02756786942604826,60.90587477250533,60.21627426147461,2.3750691033590376
+world_model,43,4.663249644366178,5.197988986968994,1.417956422526132,0.2016358456828377,0.2016417235136032,0.001053450683586269,0.07287087088281455,0.0729874148964881,0.007378596941464652,0.06596216254613613,0.064142882823944,0.004669454772836036,0.8439325040036981,0.8473471403121948,0.030309773821401938,0.8557801138270985,0.8579816222190857,0.02359073964787324,54.758696122602984,54.26784896850586,2.29666892003246
+world_model,44,4.820258552377874,5.26494026184082,1.5170198709945166,0.19141049005768512,0.1915505081415176,0.0011964629266091522,0.07463988864963698,0.0752680376172065,0.007640044738125034,0.06870337338610125,0.067132830619812,0.005154294388421187,0.8801431981000033,0.883827805519104,0.023404964688060834,0.887423499064012,0.8914567828178406,0.019321797589463268,48.52374614368785,48.17716598510742,2.130957419855159
+world_model,45,4.499387415972623,4.788283824920654,1.3419896173232597,0.18097485872832209,0.1810730099678039,0.0012238891311396635,0.07691518962383265,0.0774498432874679,0.007923242071444339,0.06993634931065816,0.069240778684616,0.005453774967110755,0.9132918444546786,0.9169639348983764,0.01700426884347143,0.9163225618275729,0.9198532104492188,0.015116529439470276,42.298390821977094,41.86894226074219,1.972031163749938
+world_model,46,4.4019012451171875,5.071197748184204,1.4210739727352106,0.17068285969170652,0.170734703540802,0.0011829686732258828,0.07766453854062337,0.0781206339597702,0.008049931328795428,0.07298356057568027,0.0725413635373115,0.005999081411846944,0.9418702288107439,0.944804310798645,0.011446851599551831,0.942465218630704,0.9449551701545716,0.010920732630230372,35.99537832086737,35.60323715209961,1.805010214627391
+world_model,47,4.580573667179454,5.063187122344971,1.4524365492833493,0.16013298386877228,0.1601799726486206,0.0012463604673350452,0.08014671910892829,0.07976645976305,0.008229790547892956,0.0767530636353926,0.0762779712677002,0.006654934270326531,0.9656780091199008,0.9678959250450134,0.006814714617034084,0.9650727618824352,0.9667105674743652,0.006947567920552799,29.655807495117188,29.381351470947266,1.595768324661803
+world_model,48,4.618325883691961,5.359715700149536,1.6106582234522935,0.14921207590536634,0.1490256339311599,0.001358504700005775,0.08153977312824938,0.080300472676754,0.008343315496332047,0.07976058938286516,0.0791859850287437,0.007278867530684533,0.983601299199191,0.9849081039428712,0.0033031324019620295,0.9828116622838107,0.9837567210197448,0.003584405022767056,22.94083161787553,22.702808380126957,1.3265387811506282
+world_model,49,4.722050731832331,5.404139995574951,1.6351713963958396,0.13684691353277723,0.1366293728351593,0.001312416491713735,0.08303084021264852,0.0814526826143264,0.008579099871448144,0.08001114157113158,0.0787350907921791,0.0075299718363697875,0.9951290813359347,0.9953976273536682,0.001001256906550283,0.9944158846681769,0.9947677850723268,0.0012460041280240354,15.543833992697976,15.287129402160645,0.9320664717399902
+world_model,50,4.503697070208463,5.038814067840576,1.588140144007771,0.1136427907781167,0.1138119250535965,0.0020374551838318717,0.10040278597311536,0.0975040048360824,0.010874890642858962,0.1058652583848346,0.1022517755627632,0.011154786421497257,0.9999999837441877,1.0,6.0144064315989e-08,0.9999999620697715,1.0,7.666893939616094e-08,0.0,0.0,0.0

unitree_g1_pack_camera/case3/backbone_profile_dense/inference/backbone_analysis/backbone_profile_report.md

@@ -0,0 +1,30 @@
+# Backbone Block Profiling Report
+
+## Dataset Overview
+
+- Rows: 27500
+- Blocks: 25
+- Steps: 1-50
+- Pass types: policy, world_model
+
+## Block Timing
+
+- `forward_time_ms`: mean=29.5324, median=33.0049, p90=53.1777
+
+## Stability
+
+- `rel_l2_delta_vs_prev`: mean=0.144351, median=0.122058, p90=0.238142
+- `cosine_vs_prev`: mean=0.986104, median=0.993058, p90=0.998209
+
+## Top Cache Candidates
+
+- `output_5` (output_blocks): tail_rel_l2=0.047653, mean_forward_time_ms=40.3060
+- `output_4` (output_blocks): tail_rel_l2=0.055735, mean_forward_time_ms=38.0808
+- `output_3` (output_blocks): tail_rel_l2=0.059179, mean_forward_time_ms=38.4340
+- `output_2` (output_blocks): tail_rel_l2=0.059520, mean_forward_time_ms=4.2037
+- `output_1` (output_blocks): tail_rel_l2=0.069876, mean_forward_time_ms=3.3372
+- `output_0` (output_blocks): tail_rel_l2=0.071423, mean_forward_time_ms=3.4297
+- `output_6` (output_blocks): tail_rel_l2=0.074268, mean_forward_time_ms=40.6642
+- `middle` (middle_block): tail_rel_l2=0.075425, mean_forward_time_ms=24.7022
+- `input_7` (input_blocks): tail_rel_l2=0.075788, mean_forward_time_ms=38.4152
+- `input_11` (input_blocks): tail_rel_l2=0.077308, mean_forward_time_ms=2.5828

unitree_g1_pack_camera/case3/backbone_profile_dense/inference/backbone_analysis/block_summary.csv

@@ -0,0 +1,26 @@
+block_stage,block_name,block_index,mean_forward_time_ms,mean_rel_l2_delta_vs_prev,mean_cosine_vs_prev,mean_l2_delta_vs_full50,mean_cosine_vs_full50,mid_mean_rel_l2_delta_vs_prev,mid_mean_cosine_vs_prev,mid_mean_forward_time_ms,tail_mean_rel_l2_delta_vs_prev,tail_mean_cosine_vs_prev,tail_mean_forward_time_ms
+output_blocks,output_5,5,40.30600001104176,0.07424595491729107,0.9978543047228196,14992.717945223721,0.8542837543921037,0.06805019055642345,0.9986008080569181,40.302541466768496,0.047652576457370366,0.9999622111970728,40.72485751133751
+output_blocks,output_4,4,38.080781668593936,0.08642426159765035,0.9958082048260436,2957.2041925881126,0.8080604154684327,0.08421756147338452,0.996300181398144,38.20055009478711,0.05573495308106591,0.9985793325034055,38.91625686294653
+output_blocks,output_3,3,38.43397183640098,0.09250370155612145,0.9952302734904918,3060.3711755371096,0.8026768178289587,0.09244359002420392,0.9955291416479911,38.81973216463348,0.0591791972179304,0.9983712329105897,38.9416516631503
+output_blocks,output_2,2,4.203701027753678,0.09137489838812475,0.9953451827859613,3606.0498737404564,0.7830393203822049,0.08808990133305386,0.995942129459216,4.195578110069682,0.059519522687928206,0.9983943538232283,4.28258259780705
+output_blocks,output_1,1,3.3372188554230062,0.10800083809236755,0.9934082725163073,1412.8108868963068,0.7369476676529104,0.107915337459543,0.9937602771825088,3.3771630963189647,0.06987590529024597,0.9975565972653302,3.3194192304191263
+output_blocks,output_0,0,3.429651934738186,0.11093722168763955,0.993079345279368,1640.254585723877,0.7372313263470476,0.11102681995663803,0.9934070663792747,3.589113329605752,0.07142288469794118,0.9974490214477886,3.3163734690540214
+output_blocks,output_6,6,40.6641822964461,0.09676676388684802,0.9951365732748566,7144.494426935369,0.8271069422635166,0.08897464989293442,0.996182621170432,41.026938741913796,0.07426767603240225,0.9976629585027694,40.46990792690353
+middle_block,middle,0,24.702236674387347,0.11546830409975349,0.992583174265382,2010.486007787531,0.7406484127044678,0.1174934316281374,0.9926540745801223,24.89245473407209,0.07542458674447099,0.9971520776098425,24.817648800936613
+input_blocks,input_7,7,38.41519514437426,0.10886899768573562,0.9936545515546993,1603.4049312106047,0.8162635827064514,0.11071356224678292,0.9936723225302511,38.47269122428321,0.07578794385560529,0.997268594936891,38.89047441030429
+input_blocks,input_11,11,2.5827501180835744,0.11662656739907049,0.9924228371074337,1722.0765785217286,0.7539892436970363,0.11981556909682942,0.9923575057353808,2.6117673756562785,0.07730753370984031,0.997005215829069,2.569907778789374
+input_blocks,input_10,10,2.72361038337377,0.11654036492800443,0.992433423401031,1607.604853695956,0.7628968893668868,0.12067004609288587,0.9922557594456198,2.6563743292656423,0.07742048667913129,0.9969992900436575,2.8803871330720456
+input_blocks,input_9,9,0.743831817314706,0.11526945793175296,0.9926486552756881,1536.432679124312,0.7748632244901223,0.11962514901251499,0.9924179907743033,0.7319237524845126,0.07781377817419437,0.9969735102220015,0.7902192989025604
+input_blocks,input_8,8,38.62152510364963,0.12054014859067924,0.9921188446276705,1432.7270319019665,0.7689293622428721,0.12640916643204622,0.9916752894203384,38.860303962611276,0.08097956512462004,0.9968675247647546,38.43976048837331
+output_blocks,output_8,8,42.76750654659488,0.1041652080542468,0.9972895507688646,13048.933399103338,0.8096005897088484,0.08988681350112988,0.9990433824784828,43.00354546020654,0.08253373083743178,1.000350392406637,42.880791404538535
+input_blocks,input_5,5,42.64259841631759,0.0933007191182399,0.9968145641745353,1739.4882183283025,0.8823152181235226,0.09050172304500745,0.9971030238386872,42.54317751609886,0.08911182880401607,0.9969788096167824,43.56268527282571
+output_blocks,output_7,7,40.16711385920644,0.11893464999484214,0.9925614916106102,4740.665635986328,0.7864092644778164,0.10539288993005624,0.9945139388243357,40.24617651223459,0.09765800244428892,0.9955882240425457,40.67105716306039
+input_blocks,input_6,6,0.595384887419641,0.13015179072743438,0.9910717374871525,1809.7961725408381,0.8091255191239444,0.12847564540384132,0.991497116687494,0.6085407196491699,0.09858578338541762,0.995211638103832,0.6194137735292315
+output_blocks,output_9,9,48.49497906059366,0.16689521762678924,0.9861997717054966,6137.993441606001,0.7435040426796133,0.14370292125197198,0.9905036210497736,48.337520912563775,0.15144165066832843,0.989499507708983,48.45326270442456
+input_blocks,input_4,4,43.2311917447739,0.1817058995504326,0.9832322480837798,2011.190616011186,0.7216005914861505,0.16583138326932856,0.9864331201299444,43.43364972548632,0.15901249613274224,0.9877330124378204,43.76259167805653
+output_blocks,output_10,10,48.60783339351077,0.19811972225143645,0.9799721469352771,4502.460150035511,0.6995518704977902,0.17081736489420843,0.9862495191169508,48.90580477850165,0.175770029425621,0.9854899059642445,48.58748410743746
+input_blocks,input_0,0,30.3599044511264,0.36612097531944127,0.9117056313134965,1307.8138972611862,0.5234270588918166,0.32535993975478333,0.9471129615069468,30.35658905665628,0.1798567053269256,0.9853965935381976,30.3359796801074
+input_blocks,input_3,3,0.6927925035018813,0.22364106857267954,0.9736635546737346,2157.1204280783913,0.6818592004342513,0.1956160919361816,0.9808968563616534,0.6883892591526627,0.19636715982447966,0.9810252604159442,0.7262287428602576
+input_blocks,input_1,1,58.22319687090137,0.25090528344510876,0.9659175354164916,1829.0903750887785,0.6146969364989887,0.2160034192266402,0.9781726463274523,58.35200590484683,0.2002870954573154,0.9814362926916642,58.548310157758266
+input_blocks,input_2,2,57.37443873701109,0.2518176541478824,0.9677503171458094,2288.248785289418,0.6070540508086031,0.21795401306121379,0.9776382349528275,57.39882213438486,0.23765106031840494,0.9729208079251376,57.36874400384047
+output_blocks,output_11,11,48.90883109235967,0.16944817667878845,0.9847041587440335,18168.413182705965,0.46862712654200467,0.1436156316527298,0.9913680543115129,48.581788756832225,0.2646951804784211,0.9647044910625978,49.03405244165862

unitree_g1_pack_camera/case3/backbone_profile_dense/inference/backbone_analysis/stage_summary.csv

@@ -0,0 +1,10 @@
+window,block_stage,mean_forward_time_ms,mean_rel_l2_delta_vs_prev,mean_cosine_vs_prev,mean_l2_delta_vs_full50,mean_cosine_vs_full50
+all,input_blocks,26.35053501482065,0.17295741061803846,0.9794528250217935,1753.749547254389,0.7264184064892205
+all,middle_block,24.702236674387347,0.11546830409975349,0.992583174265382,2010.486007787531,0.7406484127044678
+all,output_blocks,33.116814298555255,0.11815138456101215,0.9922157730550109,6784.364074673508,0.754753261520104
+mid,input_blocks,26.392852913381326,0.16141464238150463,0.9851027356425749,1838.957427945488,0.7266639219583543
+mid,middle_block,24.89245473407209,0.1174934316281374,0.9926540745801223,2007.7355251559964,0.773752135631842
+mid,output_blocks,33.21553778536967,0.10784447266052313,0.9942833950896284,6864.791992176491,0.7592755320816587
+tail,input_blocks,26.541225201534953,0.12918178639939107,0.9904847125438127,880.7333542390303,0.9334695872032281
+tail,middle_block,24.817648800936613,0.07542458674447099,0.9971520776098425,731.289155925404,0.9663700651038777
+tail,output_blocks,33.29980809022813,0.10081260910991462,0.9936340190695994,3178.3527087298307,0.9287726470466816

unitree_g1_pack_camera/case3/backbone_profile_dense/inference/round_summary.csv

@@ -0,0 +1,12 @@
+sample_id,scene,round_id,policy_pass_total_time_s,world_model_pass_total_time_s,round_total_time_s,latent_init_dist_to_prev_round,action_drift_vs_prev_round,psnr_full50
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,0,272.30813097953796,258.77083349227905,625.7526919841766,,,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,1,186.9196274280548,263.2243719100952,540.1022210121155,221.82896423339844,0.33534055948257446,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,2,244.72048449516296,254.23945784568787,588.8439288139343,235.9840545654297,0.1256718635559082,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,3,263.5875928401947,254.7561550140381,607.7824947834015,199.4680633544922,0.18907040357589722,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,4,217.18552041053772,256.91192841529846,562.6434659957886,260.11322021484375,0.6156945526599884,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,5,252.68572807312012,261.74088191986084,602.5384485721588,190.06353759765625,0.43064773082733154,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,6,255.90784645080566,249.05019426345825,593.9222755432129,228.24305725097656,0.09470903873443604,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,7,226.4355342388153,271.73611545562744,585.1254336833954,228.67453002929688,0.14335262775421143,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,8,269.48164653778076,265.67703223228455,625.1090137958527,250.22793579101562,0.6966221630573273,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,9,259.6198160648346,210.60337662696838,558.2484686374664,248.954345703125,0.14137792587280273,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,10,94.52964091300964,90.96858811378479,265.539076089859,206.26715087890625,0.03912711143493652,

unitree_g1_pack_camera/case3/backbone_profile_dense/inference/sample_summary.csv

@@ -0,0 +1,3 @@
+sample_id,scene,pass_type,pass_total_time_s,action_first_stable_step,state_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_state,oracle_budget_latent,latent_init_dist_to_prev_round,action_drift_vs_prev_round,round_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,psnr_full50
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,231.21650622107765,,,,44.90909090909091,46.72727272727273,46.72727272727273,46.54545454545455,,226.98248596191405,0.2811613976955414,559.6006835373965,231.21650622107765,239.78899411721662,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,239.78899411721662,,,,45.18181818181818,46.90909090909091,46.90909090909091,46.81818181818182,,226.98248596191405,0.2811613976955414,559.6006835373965,231.21650622107765,239.78899411721662,

unitree_g1_pack_camera/case3/output/inference/analysis/analysis_report.md

@@ -0,0 +1,54 @@
+# DDIM Analysis Report
+
+## Analysis Config
+
+- Stability threshold: None
+- Stability window: 3
+- Mid-step window: 15-35
+
+## Dataset Overview
+
+- Unique samples: 1
+- Step rows: 1100
+- Sample rows: 2
+- Round rows: 11
+- Scenes: unitree_g1_pack_camera
+- Pass types: policy, world_model
+
+## Five Directions
+
+### Original Early Stop
+
+- `latent_first_stable_step_policy`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_first_stable_step_world_model`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_l2_vs_full50`: mean=116.0635, median=135.1986, p90=161.0852, p95=169.8322
+
+### Direction C: Action Converges First
+
+- `action_first_stable_step`: mean=nan, median=nan, p90=nan, p95=nan
+- `latent_first_stable_step`: mean=nan, median=nan, p90=nan, p95=nan
+- `action_vs_full50_95pct_step`: mean=46.6364, median=46.6364, p90=46.6364, p95=46.6364
+- `share_action_first_stable_before_latent`: nan
+- `share_action_95pct_before_latent`: nan
+
+### Direction D: Cross-Step Similarity
+
+- `latent_delta_mid_steps`: mean=0.3529, median=0.3489, p90=0.4169, p95=0.4276
+- `action_delta_mid_steps`: mean=0.0320, median=0.0307, p90=0.0455, p95=0.0488
+- `state_delta_mid_steps`: mean=0.0312, median=0.0301, p90=0.0439, p95=0.0465
+
+### Direction A: Budget Heterogeneity
+
+- `oracle_budget_action`: mean=46.6364, median=46.6364, p90=46.6364, p95=46.6364
+- `oracle_budget_state`: mean=46.7273, median=46.7273, p90=46.7273, p95=46.7273
+- `oracle_budget_action_by_scene`:
+  - `unitree_g1_pack_camera`: mean=46.6364, median=46.6364, std=0.0000
+
+### Direction B: Round Reuse
+
+- `latent_init_dist_to_prev_round`: mean=215.2431, median=230.0003, p90=237.2570, p95=239.0189
+- `action_drift_vs_prev_round`: mean=0.2697, median=0.2050, p90=0.4657, p95=0.6923
+- `round_total_time_s`: mean=147.1026, median=152.2600, p90=152.9385, p95=152.9569
+- `policy_pass_total_time_s`: mean=71.2595, median=74.7608, p90=74.9093, p95=75.1472
+- `world_model_pass_total_time_s`: mean=73.1823, median=74.7396, p90=74.9509, p95=74.9599
+

unitree_g1_pack_camera/case3/output/inference/analysis/scene_summary.csv

@@ -0,0 +1,4 @@
+scene,pass_type,pass_total_time_s,pass_total_time_s,pass_total_time_s,action_first_stable_step,action_first_stable_step,action_first_stable_step,state_first_stable_step,state_first_stable_step,state_first_stable_step,latent_first_stable_step,latent_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_90pct_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,action_vs_full50_95pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_action,oracle_budget_action,oracle_budget_state,oracle_budget_state,oracle_budget_state,round_total_time_s,round_total_time_s,round_total_time_s,policy_pass_total_time_s,policy_pass_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,world_model_pass_total_time_s,world_model_pass_total_time_s,latent_init_dist_to_prev_round,latent_init_dist_to_prev_round,latent_init_dist_to_prev_round,action_drift_vs_prev_round,action_drift_vs_prev_round,action_drift_vs_prev_round
+,,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std
+unitree_g1_pack_camera,policy,71.25950481674887,71.25950481674887,,,,,,,,,,,44.72727272727273,44.72727272727273,,46.63636363636363,46.63636363636363,,46.63636363636363,46.63636363636363,,46.72727272727273,46.72727272727273,,147.10261366584083,147.10261366584083,,71.25950481674887,71.25950481674887,,73.18230964920737,73.18230964920737,,215.24309921264648,215.24309921264648,,0.2696936547756195,0.2696936547756195,
+unitree_g1_pack_camera,world_model,73.18230964920737,73.18230964920737,,,,,,,,,,,45.0,45.0,,46.72727272727273,46.72727272727273,,46.72727272727273,46.72727272727273,,46.72727272727273,46.72727272727273,,147.10261366584083,147.10261366584083,,71.25950481674887,71.25950481674887,,73.18230964920737,73.18230964920737,,215.24309921264648,215.24309921264648,,0.2696936547756195,0.2696936547756195,

unitree_g1_pack_camera/case3/output/inference/analysis/step_aggregate.csv

@@ -0,0 +1,102 @@
+pass_type,step,step_time_s,step_time_s,step_time_s,latent_delta,latent_delta,latent_delta,action_delta,action_delta,action_delta,state_delta,state_delta,state_delta,action_cosine_vs_full50,action_cosine_vs_full50,action_cosine_vs_full50,state_cosine_vs_full50,state_cosine_vs_full50,state_cosine_vs_full50,latent_l2_vs_full50,latent_l2_vs_full50,latent_l2_vs_full50
+,,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std,mean,median,std
+policy,1,1.4199446548115124,1.4689834117889404,0.24784095387844252,1.4148756699128584,1.4147228002548218,0.0017983203653821737,0.04296489059925074,0.0397998988628387,0.009299621251750496,0.03482877835631364,0.0343807376921176,0.00547841440970538,0.06929501908069303,0.0869507640600204,0.08247889065490668,0.07147780297832051,0.0868794545531272,0.07320887474281128,166.28865189985797,165.389892578125,10.731242910725213
+policy,2,1.3961981426585803,1.4666833877563477,0.23087492474052587,1.0217565189708362,1.0221338272094729,0.0018247056445979686,0.03199359351261092,0.0317323058843612,0.0037646796769319367,0.025336772880770908,0.0253736060112714,0.003678443038993123,0.06810338879850776,0.0875042751431465,0.08210616837812114,0.07210849614983253,0.0864711329340934,0.07331465993546059,166.4136102849787,164.7098846435547,11.000094977145157
+policy,3,1.3912982073697178,1.4673566818237305,0.22973540777621523,0.8505897142670371,0.8505821228027344,0.0010051918118005416,0.024692537601698473,0.0248304959386587,0.0038281192796053333,0.019629743881523547,0.01930976472795,0.0036561174108232456,0.07052726840431038,0.0910862684249877,0.08170359839211622,0.07576540315693073,0.0877224504947662,0.07393231993938981,165.5298420299183,163.74937438964844,10.924222893091711
+policy,4,1.3955245884982022,1.4692416191101074,0.23115731906112152,0.7504698092287237,0.7511569857597351,0.002213468830741041,0.0209008015014908,0.0201370678842067,0.004417066023597906,0.01636863525279539,0.0147214485332369,0.004398287368375191,0.07521781071343202,0.0970934405922889,0.08137898664731064,0.08186778341504657,0.0920410752296447,0.0744835682426777,165.39599748091265,162.8660888671875,10.500462465080362
+policy,5,1.3938182484019885,1.4660944938659668,0.2294866148731624,0.682613958011974,0.6820001602172852,0.0018772849971748747,0.017487539825114282,0.0166238695383071,0.0033519486171890347,0.013972549783912526,0.0127493571490049,0.0031853362784936253,0.0814702175557613,0.1048394218087196,0.08113645505303767,0.08942560546777466,0.098782017827034,0.07517409688283853,164.2550714666193,163.5480194091797,11.22196545262611
+policy,6,1.3910709077661687,1.467343807220459,0.22875892281431426,0.6330291791395708,0.6330946683883667,0.0013312794694004318,0.016232270154763292,0.0151452394202351,0.003113519461051901,0.012189679017121091,0.0117673017084598,0.0014661549227332263,0.08859342505986037,0.1138118654489517,0.0809000431229695,0.09813935368914492,0.1066522896289825,0.07590897326470579,162.82386363636363,162.26492309570312,9.8885045156412
+policy,7,1.3885809941725298,1.467557430267334,0.22863482631318804,0.5942280996929515,0.5937097668647766,0.0011322756789933666,0.0162948049943555,0.0147878648713231,0.0031789977673532957,0.012443589385260183,0.0120103694498538,0.0015854616989472731,0.0963388718664646,0.1232869997620582,0.08074121432615688,0.10775288592346687,0.1153533160686492,0.07666288371092304,159.67699501731178,160.8734130859375,8.891887204756257
+policy,8,1.3973513083024458,1.4665753841400146,0.23078928445988792,0.5634413686665621,0.5633506774902344,0.0015484535584969308,0.01621390616690567,0.0149094723165035,0.0030156220036154,0.013134431601925291,0.0125940218567848,0.0016045019236136716,0.10475609265267845,0.1333369612693786,0.08063809382102032,0.11806750229813832,0.1247528344392776,0.07738447291690853,158.36621509898794,159.92706298828125,8.676093227119262
+policy,9,1.3959431431510232,1.467972755432129,0.2328118621935322,0.5371802340854298,0.5368256568908691,0.0011949182832252372,0.015963013622571073,0.0149467037990689,0.002758452240668138,0.013670662248676408,0.0131606245413422,0.0016642485039808598,0.11386911875822324,0.1439066082239151,0.0804174499320011,0.12889386645772236,0.1347137689590454,0.07806687096551898,157.12624983354047,157.07748413085938,8.89944298073369
+policy,10,1.39333776994185,1.466566801071167,0.23291327214493826,0.5157071514563127,0.5156304240226746,0.00071095651044754,0.0161112701858986,0.0151212671771645,0.002551207776720993,0.014149567281657937,0.0134865520521998,0.0016877545930272305,0.12376608089967203,0.1559549868106842,0.08014399461680466,0.1402043910189108,0.1449289619922638,0.07874248665891999,154.459340875799,154.81410217285156,8.866539908645626
+policy,11,1.3970889178189365,1.4682331085205078,0.23354156079418062,0.4965983602133664,0.496254950761795,0.0013183721506610222,0.0162570081143216,0.0154960826039314,0.0022417798275051622,0.015134023587134691,0.0145450728014111,0.0015748084205547395,0.1344618482345884,0.1693732738494873,0.07985499531821201,0.15204009752381928,0.1559115648269653,0.07947249003916608,152.2045246471058,152.67138671875,8.152501028924327
+policy,12,1.39147517897866,1.4630882740020752,0.23190521158086197,0.480160038579594,0.480011373758316,0.0008933944401713886,0.0167662082239985,0.016273895278573,0.0021683699879489573,0.01598141748796804,0.0154201798141002,0.0014003313754610057,0.14549739370969206,0.1823156476020813,0.07952362047116966,0.1642640690234574,0.1667821109294891,0.08012449297972861,150.1994295987216,147.66102600097656,8.361983984491255
+policy,13,1.3986355391415684,1.4684324264526367,0.22967293422642662,0.46388220245187933,0.4637138843536377,0.0009076149443483036,0.017447877844626206,0.0169204361736774,0.002141879203330685,0.016630208966406865,0.0162317547947168,0.0014006995934814526,0.156927143138918,0.1956851184368133,0.07940099614584072,0.17687058245593848,0.177789956331253,0.08076975496439211,148.61978149414062,146.8197784423828,7.4275339076589475
+policy,14,1.395016843622381,1.4685349464416504,0.23357109532484116,0.4508358565243808,0.450634241104126,0.0008967817091598962,0.01837730119851497,0.017890753224492,0.002275769719796006,0.017592312683435953,0.0173444207757711,0.001335839703994519,0.16876144300807605,0.209689050912857,0.07932480164353904,0.1899092434482141,0.1891032308340072,0.0813762789687884,147.01048001376066,145.48289489746094,7.128993719840749
+policy,15,1.3874100988561457,1.4646108150482178,0.23205342475211724,0.43886817314408044,0.438880980014801,0.0005999638861373147,0.019272385165095253,0.0187625139951705,0.002319646962380556,0.018260731277140663,0.0180140379816293,0.0013289654973574485,0.1810012679039077,0.2238913625478744,0.07920025056314642,0.2032911188223145,0.2009234279394149,0.08197090542939899,145.66456326571378,144.00595092773438,6.468737494900389
+policy,16,1.3968654112382368,1.4682765007019043,0.2334989595694358,0.4266154440966519,0.4266196191310882,0.0006458059292767522,0.020239090377634182,0.0196974910795688,0.0023352447714596316,0.0193730694326487,0.0192009173333644,0.0012392080848190256,0.19371072524650526,0.237640231847763,0.07902959158847749,0.2171774411743337,0.2154812514781952,0.08246736938135521,144.02324052290484,142.64797973632812,5.964588441657
+policy,17,1.395165746862238,1.467644453048706,0.23266883953622636,0.4158513464710929,0.4163050353527069,0.0009880957270411537,0.02115450240671629,0.0206316318362951,0.0024071820959970154,0.020321888341145025,0.0203274842351675,0.0011518107421581315,0.20679082399742166,0.2509100139141083,0.07888301141397043,0.23139992153102698,0.2292900681495666,0.08287899364964434,142.933369029652,141.83502197265625,5.620575124261225
+policy,18,1.397233334454623,1.4682981967926023,0.23322359019666616,0.4061706851829182,0.4059041142463684,0.0010675779126346565,0.021897767077792736,0.0213636942207813,0.0022499933095869906,0.02109132707118983,0.0212969668209552,0.0011487422475652763,0.22054141522808504,0.264401227235794,0.0785308469748189,0.2459408884698694,0.2424930930137634,0.0832141244080728,142.01401034268466,140.7686309814453,5.277748181651768
+policy,19,1.3941774801774458,1.468799114227295,0.2332288031057206,0.3966258032755418,0.3966047167778015,0.0009748517740412879,0.02329360321164126,0.022596351802349,0.002533012080381376,0.022247372533787335,0.0226513762027025,0.0011441157690688817,0.2342805740508166,0.2779521048069,0.07833726484104021,0.26109283756126056,0.2563294470310211,0.08350101400361967,141.3571250221946,140.20330810546875,4.899095781518582
+policy,20,1.39013671875,1.4671165943145752,0.23162481996733364,0.38738407871939917,0.3872357010841369,0.000899909016559721,0.024091059003363915,0.0234279241412878,0.0024521910605300157,0.023034107109362382,0.02347251214087,0.0011936538329097878,0.24874622781168326,0.2917625308036804,0.07789532063663962,0.27672880887985224,0.2705399394035339,0.08387115602876005,140.51580533114347,139.3236541748047,4.806872258645565
+policy,21,1.3930827487598767,1.468642234802246,0.23267148655937464,0.3787874471057545,0.3791123330593109,0.0010040153324427045,0.025292632254687163,0.0245395693928003,0.002548339825735864,0.02440367469733407,0.0248856451362371,0.0011767432557690904,0.26387040181593463,0.3060960173606872,0.07716367569443724,0.29291873086582526,0.285388708114624,0.08409135178697852,139.62204534357244,138.72824096679688,4.422310312428542
+policy,22,1.3980731530623003,1.4692797660827637,0.23324336723607852,0.37090576778758655,0.3710913062095642,0.0006050172995983032,0.02652050300755278,0.0259991902858018,0.00266291533362873,0.025566521524028298,0.0258515384048223,0.0012256957443174863,0.2793912481177937,0.3211789429187774,0.07663848245315401,0.30966383760625665,0.3006980121135711,0.08423221750470727,138.8831551291726,137.80862426757812,4.253026240154828
+policy,23,1.3876693465492942,1.4661476612091064,0.2321090896186312,0.3632450808178294,0.3631476759910583,0.0009504257382545858,0.02797194892032574,0.0275118649005889,0.002786243580084514,0.02672967907379969,0.026794483885169,0.0013123951879026217,0.29518670656464313,0.3364537358283996,0.07631225766234985,0.3269407559524882,0.3165871500968933,0.08421111924698894,138.0939580743963,136.85728454589844,4.05977729850141
+policy,24,1.3870555270801892,1.4667108058929443,0.23229109916023472,0.356257368217815,0.3562216162681579,0.0008385704575403215,0.028949941931800352,0.0283108800649642,0.0025630100500371377,0.0280379246581684,0.0278217382729053,0.0013886491431190028,0.3124776509675113,0.3526518940925598,0.07564842478126434,0.34491849216547876,0.3331633508205414,0.08416368966062525,137.34270130504262,136.33912658691406,3.824452310262525
+policy,25,1.3811318657614968,1.4669959545135498,0.23213730278922332,0.3493820916522633,0.3490746021270752,0.0008195303506146058,0.03037995083088219,0.0296198800206184,0.0026112004147170534,0.029447816650975785,0.0291819125413894,0.0014741240479695186,0.33067775856364856,0.3696516454219818,0.07493827461481974,0.36359960382634937,0.3542846441268921,0.0840333271110926,136.6431288285689,136.1310577392578,3.4707142732339618
+policy,26,1.397722070867365,1.468933343887329,0.23269373108215585,0.34216034141453827,0.3419646620750427,0.000906596112744071,0.03188043527982448,0.031423483043909,0.0025717996052750554,0.030636898495934164,0.0305297560989856,0.0015972289996493473,0.3495491417971524,0.387301504611969,0.07412730112239446,0.3829627199606461,0.3768246173858642,0.08378331506297067,136.12241987748578,135.1168975830078,3.228142836090027
+policy,27,1.3976352431557395,1.4682977199554443,0.23002463328961734,0.33548542857170105,0.3353995084762573,0.001100893883873762,0.03336546654728322,0.0329380482435226,0.0024973943812357903,0.032164451073516453,0.0323128663003444,0.0017110141330944424,0.36968166719783435,0.4060558676719665,0.0730845134785985,0.40315917947075586,0.4002630710601806,0.08345396989649267,135.60153475674716,134.41632080078125,3.1084850058021094
+policy,28,1.3889400092038242,1.4663820266723633,0.23109903297095374,0.32923729311336164,0.3289990723133087,0.0009212966334163451,0.03477015380154951,0.034459076821804,0.0023810203645046597,0.033644755286249196,0.0338968634605407,0.0018299421679751377,0.39076151902025397,0.4258028268814087,0.07204199367493073,0.4241756254976446,0.4218252599239349,0.08301130653093475,135.10802112926137,133.9518585205078,2.944620540586098
+policy,29,1.3958456733010032,1.4680793285369873,0.23273762887265056,0.3229081793264909,0.3226611912250519,0.0010644529768016925,0.03650624169544736,0.0360220894217491,0.002548195799598,0.03528034043582999,0.0353871285915374,0.0018673795680370967,0.41337809237566864,0.4466045200824737,0.07058659685275215,0.44607106663964013,0.4431875646114349,0.08238279481466829,134.4320331920277,133.38516235351562,2.8918477732866448
+policy,30,1.398771047592163,1.4687495231628418,0.23035881097902663,0.3166154704310677,0.3168188631534576,0.0009145729568201159,0.038377677852457166,0.0380931757390499,0.0027605747591740886,0.03727676279165523,0.0372313372790813,0.0020546558874363053,0.43668326464566315,0.4683857858181,0.06917588424982536,0.4689953787760301,0.4654948115348816,0.08148607605333388,134.1102724942294,133.15280151367188,2.7293462311716934
+policy,31,1.3975862156261096,1.468428134918213,0.22974310788986158,0.31156241893768305,0.3115926086902618,0.0008053260474167256,0.04030895876613525,0.0395881086587905,0.002983343898059974,0.039029992778192815,0.0389085449278354,0.0022026435383434026,0.4615032970905304,0.491204559803009,0.06745016506014705,0.4928771636702798,0.488850474357605,0.08034385390111914,127.84154371781783,126.80603790283205,2.684130126755344
+policy,32,1.398115028034557,1.4685206413269043,0.23000187500011807,0.3038616180419922,0.3036515414714813,0.0008707003144547423,0.04247550835663617,0.0416393354535102,0.0030729552467033874,0.04031326147642998,0.0403376780450344,0.002325486794508347,0.4871742996302518,0.515261173248291,0.06576557929649716,0.5176525278524919,0.5128324031829834,0.07914143450948494,121.61492087624289,120.6345443725586,2.6838542257147173
+policy,33,1.3982429721138694,1.4688405990600586,0.23005209178713826,0.29566746137358924,0.2956995069980621,0.0008574776753279273,0.04434451528570864,0.0434012822806835,0.003158340661104034,0.042377971790053585,0.0424748510122299,0.0025846454140460735,0.515161007642746,0.5403033494949341,0.06306635073559665,0.5436133335937153,0.5380820035934448,0.07763774930639691,115.44529099897905,114.56948852539062,2.597724000539393
+policy,34,1.3965325572273948,1.46874737739563,0.22887058813830843,0.28751979903741315,0.2876001894474029,0.0008464997297958268,0.046543978493321994,0.045653935521841,0.003269844393746624,0.044370111755349384,0.0446448698639869,0.00285659600654204,0.5442631082101301,0.5665419101715088,0.0602054745798515,0.570698781446977,0.5645166039466858,0.0757948558162599,109.20886091752486,108.38751983642578,2.551376837960185
+policy,35,1.3976473158056086,1.468266487121582,0.22980796371659562,0.27876275235956366,0.2786843478679657,0.0008252766096163018,0.04889900508252052,0.0482247583568096,0.003459421620781423,0.046336260031570035,0.0467241965234279,0.0031529949394377126,0.574638158082962,0.5942624807357788,0.057110933262157236,0.598864807323976,0.5921886563301086,0.07356039683880124,103.00947432084517,102.27365112304688,2.461374086178861
+policy,36,1.3855494369160046,1.4689552783966064,0.22718398608558008,0.2700976431369781,0.269994705915451,0.0008440204694016255,0.05155899951403788,0.0506390631198883,0.0037805401479951094,0.04937205565246665,0.0498123988509178,0.0036419656105230675,0.606609740040519,0.6233117580413818,0.053493952247218146,0.6285896680571816,0.6216908097267151,0.07061005208978514,96.91611272638494,96.29117584228516,2.463765083646315
+policy,37,1.3926266106692227,1.4650709629058838,0.22958450670246977,0.26092292774807324,0.2607474625110626,0.0009192696942363682,0.054065613245422146,0.0523351542651653,0.004276431205171295,0.05044958367943759,0.05068364366889,0.0037977112872810363,0.640471409667622,0.6533941626548767,0.0491285779000502,0.6588581746274774,0.6515571475028992,0.06759578421173414,90.7248798717152,90.28363037109376,2.4495879416903743
+policy,38,1.3983956250277432,1.4685688018798828,0.2326188678120393,0.2516408589753238,0.2514135837554931,0.0008444321745491667,0.05646827579899263,0.0541640482842922,0.004678404861550862,0.05393095788630567,0.054175067692995,0.004422981602228341,0.6756270202723417,0.6845731139183044,0.04418564613377763,0.6906227523630316,0.685060441493988,0.06362939068967274,84.54249988902698,84.22433471679688,2.3823459159131297
+policy,39,1.3974752209403298,1.4687211513519287,0.23283867121485227,0.24206021563573313,0.2421729564666748,0.0009790923877469814,0.059121415696360785,0.0564726628363132,0.005027217049753611,0.055872542614286556,0.0560077875852584,0.004829634823848823,0.7119568152861162,0.7167223691940308,0.0388637745502593,0.7230992696501992,0.7193184494972229,0.05931103042289549,78.49464416503906,78.10062408447266,2.3544623119816515
+policy,40,1.3975018587979404,1.4686484336853027,0.23203024824890445,0.23246509920467026,0.232818603515625,0.0010950287788950672,0.06186196749860586,0.0588464587926864,0.005349256306551369,0.056991226293823895,0.0569765381515026,0.005178340071205947,0.748980088667436,0.750083863735199,0.03335648708266333,0.7555004412477667,0.7530468106269836,0.054697207446160793,72.35493816028942,71.98544311523438,2.317209349338933
+policy,41,1.3991471854123203,1.468951940536499,0.23291801789241035,0.22204172882166773,0.2224677354097366,0.001190453320234576,0.06633509898727584,0.0637814998626709,0.005403513605827331,0.06074961715123866,0.0606111101806163,0.005962467886783079,0.7867171547629617,0.7849864959716797,0.028110372840959918,0.7888143712824042,0.7875792980194092,0.04906218150994738,66.28471651944247,65.92801666259766,2.2235564094492397
+policy,42,1.395815285769376,1.4650049209594729,0.2315630238006855,0.21240549602291797,0.2128109633922577,0.0011945403969590655,0.06771188974380488,0.0657639876008033,0.005658840986010276,0.06272197040644555,0.0624122582376003,0.006480972225632828,0.823605862530795,0.8200478553771973,0.022975973263899555,0.8218092755837874,0.8213566541671753,0.043080852626129006,60.175458734685726,59.91341400146485,2.090535790869781
+policy,43,1.3980659571560947,1.4691872596740725,0.23250785714120023,0.20192084935578428,0.2022386640310287,0.0011738066056845883,0.06995729627934365,0.0691783428192138,0.005586429083625461,0.06541585549712177,0.0650460347533226,0.007163909521647139,0.85914761911739,0.8551719188690186,0.01825310665284656,0.8542784723368558,0.8543077111244202,0.03662391680419965,54.12357156926935,53.80431365966797,2.0296692321248337
+policy,44,1.396042607047341,1.4662487506866455,0.23174950990005813,0.19176711954853748,0.1921779662370681,0.001051914891974807,0.07136923413385039,0.0713838636875152,0.005440439701368714,0.06831177865917028,0.0676725655794143,0.008002730695230581,0.8922106677835638,0.8895732760429382,0.014029720634793148,0.8856175270947543,0.885894775390625,0.029687563916090214,48.000628384676844,47.77872085571289,1.9226493722774975
+policy,45,1.3873529867692427,1.46867036819458,0.2318101356731532,0.1814325167374177,0.1817986816167831,0.001362562766737784,0.07331087101589544,0.0738725811243057,0.0053724983767494605,0.06970266083424737,0.068978376686573,0.008414717337615875,0.9222341125661676,0.9203423261642456,0.010291981780351706,0.9145046364177357,0.9147323369979858,0.02292448004186011,41.84291076660156,41.73094940185547,1.8037654835085017
+policy,46,1.3947387175126509,1.46449875831604,0.2315420883783965,0.17103424532846964,0.1712109446525573,0.0011189730727775157,0.07392106069759884,0.0751156657934188,0.005324450286171255,0.07290775904601265,0.0723533928394317,0.00927527095650564,0.9479800787839022,0.946979820728302,0.007117330144233319,0.9408404772931879,0.9409552216529846,0.016264008119090536,35.59961145574396,35.4770622253418,1.6463011656811974
+policy,47,1.3933005332946777,1.4662511348724363,0.23068317654106377,0.1605424054644324,0.1608028709888458,0.0013902221699989997,0.07609803026372731,0.0770305842161178,0.0052345108544724185,0.07695857029069551,0.0765223056077957,0.010292512062775956,0.9693101102655585,0.9689814448356628,0.004439686919034507,0.9638550985943187,0.963843822479248,0.010069742829094211,29.28877657110041,29.279510498046875,1.4706735852626311
+policy,48,1.397290663285689,1.4673364162445068,0.2321224198821918,0.14959971200336106,0.1498794555664062,0.0012506429864215741,0.07728927108374505,0.0782783478498458,0.005352537995597017,0.08034498786384403,0.0799207463860511,0.011077113740668124,0.985328267921101,0.9852620363235474,0.002274625257182432,0.9821105382659219,0.9819733500480652,0.004997827129811236,22.673615022139117,22.67072296142578,1.270955483369366
+policy,49,1.3973293521187522,1.467585802078247,0.23217040990034032,0.13729143007235092,0.1376739293336868,0.0013042199426100993,0.07874605940146875,0.0802952870726585,0.005839733080200932,0.08107475326819846,0.0812594220042228,0.010867235661918134,0.9956435019319708,0.9956248998641968,0.0007228876237728988,0.9941713051362471,0.9940559267997742,0.0016669520307079108,15.34714412689209,15.306425094604492,0.9137884039077372
+policy,50,1.3971644098108464,1.4685523509979248,0.23251888924898376,0.11407934467900879,0.1141030043363571,0.0020744413094250223,0.09504244747486977,0.0951025709509849,0.00769284831398244,0.10774176838723094,0.1095263659954071,0.015006327893710066,0.9999999729069796,0.9999999403953552,8.985738296640884e-08,0.9999999945813959,1.0,7.750687617115007e-08,0.0,0.0,0.0
+world_model,1,1.397060134194114,1.468520164489746,0.23209641671372636,1.4132862849669023,1.4142225980758667,0.0032558895680461536,0.04229140010747038,0.0409311316907405,0.0048616917218485,0.030018669807098067,0.0310924090445041,0.0031556253909547083,0.021687547676265243,0.0026827091351151,0.10126472276664031,0.008256774734367018,0.0062365746125578,0.0445626736590254,166.02682772549716,162.5775604248047,15.419021785677117
+world_model,2,1.394958647814664,1.4676237106323242,0.23207357875643245,1.020930290222168,1.0212992429733276,0.0015413107354754262,0.03220456330613652,0.0317432433366775,0.0038627263754493946,0.022725183516740775,0.0215565040707588,0.0021731112312203607,0.020730182613161494,0.0016211019828915,0.10121049942053982,0.00953977072442118,0.0084504354745149,0.044105362033503764,164.77466097745028,161.25946044921875,14.467070278368181
+world_model,3,1.3934266567230225,1.4684956073760986,0.23111879114398137,0.8507108363238248,0.8503804802894592,0.001811010727927916,0.02495878240601579,0.0242328494787216,0.0036585303646759566,0.017795011655173474,0.0171933639794588,0.0020872146391280122,0.02317974140698258,0.0055888779461383,0.10058269802793683,0.014085429805246257,0.0125316502526402,0.044229021902067275,164.28697204589844,160.899658203125,15.48904511224389
+world_model,4,1.3979400287974963,1.4683723449707031,0.23246778910080879,0.7501540834253485,0.7506182193756104,0.00191794380486534,0.021089591932567636,0.0213051158934831,0.0029972288205221175,0.01570766038176684,0.0147627918049693,0.002690648511058169,0.02762934311547061,0.0120556820183992,0.09982644118346136,0.02095097282224081,0.0191869363188743,0.04430270239286057,163.28567227450284,161.07655334472656,15.415127358962808
+world_model,5,1.3983885374936191,1.468574047088623,0.23280278172447721,0.6832515543157404,0.6833576560020447,0.001681939813460379,0.018667803653939163,0.0181663390249013,0.0025819645954901827,0.015457165258174555,0.0141863077878952,0.003538232208640517,0.03337351673028686,0.0196057111024856,0.09887086836198027,0.029003180097788548,0.0277245230972766,0.04449561233776792,162.13086769797584,160.75125122070312,13.982188192131357
+world_model,6,1.3960540077903054,1.4680912494659424,0.2316625904965525,0.6333704374053262,0.6337016224861145,0.002215249785508791,0.01739421249790621,0.0166730657219886,0.002509113861109463,0.015047600662166372,0.0138655416667461,0.0032147458344210277,0.03995440409264782,0.02770771458745,0.09791121376330769,0.037967261180958944,0.037167489528656,0.044735434555019854,161.13878146084872,156.13136291503906,14.202085207291068
+world_model,7,1.3958736332980068,1.4679224491119385,0.23210035252771166,0.5945573557506908,0.5951321721076965,0.00173232065280563,0.017180242067710873,0.016549339517951,0.0024768068562493637,0.014550924131816035,0.0138543220236897,0.0022567814601639814,0.047073411162603974,0.036384891718626,0.09718055479183327,0.047818138687448046,0.0476803258061409,0.04504644132232273,158.64647605202416,156.3931884765625,12.362258108487255
+world_model,8,1.3991979902440852,1.469386339187622,0.2319427273547264,0.5636275410652161,0.5633426308631897,0.0017189724443863205,0.017107021740891674,0.0165111795067787,0.002557357741843461,0.01465057039802718,0.0146069368347525,0.0013285622527370372,0.05490768692371518,0.0457929074764251,0.09668817116632726,0.058358248831196224,0.0591987632215023,0.0454010649943772,156.91902992942116,155.96109008789062,10.779268567347303
+world_model,9,1.391722484068437,1.4691870212554932,0.2292507418120153,0.5375344157218933,0.5376505851745605,0.0022545475359390396,0.017001958563923784,0.0166283417493104,0.0024255624012909176,0.015165783295577183,0.0152348224073648,0.0018196034270450908,0.06346193286166946,0.0556991249322891,0.09623736242991819,0.06950334967537358,0.071213848888874,0.045829730841848854,155.01025668057528,152.654052734375,10.49791551508185
+world_model,10,1.3981911702589556,1.4683139324188232,0.23248787471553783,0.5162937912074003,0.5161793828010559,0.0010167763264627432,0.01720741433514786,0.0167896971106529,0.002176760822726805,0.015809652734209156,0.015449957922101,0.0022760748159761177,0.07281269437887451,0.0658258944749832,0.09575691815358853,0.0811116681857542,0.0821476653218269,0.046397426351618566,152.84031538529828,150.84231567382812,9.964161531377957
+world_model,11,1.3956788236444646,1.4688832759857178,0.23241543322742958,0.49643538756804034,0.4964897930622101,0.0005655437039052367,0.017352687448940435,0.017376510426402,0.0019209104843523575,0.016932337121529983,0.0164513736963272,0.0023384868502260205,0.08292715962637552,0.0768075138330459,0.09518712940327317,0.09322815320708533,0.0936360359191894,0.046954418976113126,151.07225036621094,149.58299255371094,10.010136041633027
+world_model,12,1.3990216905420476,1.4697089195251465,0.2322847346749956,0.4789709746837616,0.4791893362998962,0.0013803605387985585,0.017910233190791144,0.0179083868861198,0.0018252727313546205,0.01752979308366771,0.0175561103969812,0.0013328470336290599,0.09345635450021784,0.0879065468907356,0.09433913289498083,0.10586790740489956,0.1052672937512397,0.04745464171535585,149.69276150790128,148.0255126953125,9.18280932526007
+world_model,13,1.390719717199152,1.4685919284820557,0.20946664553277763,0.4638652205467224,0.4639376103878021,0.0009622472656286709,0.018491224470463617,0.0182679519057273,0.00197564435588444,0.017889904027635347,0.0177722815424203,0.0011152153802464436,0.10441743650219654,0.0993654280900955,0.09361222908455707,0.11907447197220536,0.1173175722360611,0.04781402963541712,148.1121035489169,146.2946014404297,8.690717804514064
+world_model,14,1.4389164664528586,1.4685285091400146,0.07329649664424152,0.45071856542067096,0.4508586525917053,0.001068924655483121,0.019590353423898817,0.0192439910024404,0.002048634945749469,0.01884764602238478,0.0188638139516115,0.0010585285250784926,0.11569952168925238,0.1104635149240493,0.09276302399319082,0.13285514373670923,0.1303670704364776,0.048212421446734315,146.06091169877485,145.39466857910156,8.242440226012635
+world_model,15,1.4515010443600742,1.4686095714569092,0.05157234523099491,0.43811914595690643,0.4380188286304474,0.0009252998616423729,0.020062531090595446,0.0196704994887113,0.001957660520956597,0.01943665455010798,0.0194491948932409,0.0009488555749195263,0.12765153361992398,0.1225744932889938,0.09194315908723384,0.14713256061077115,0.1446343809366226,0.04855355774712764,145.3772638494318,145.15838623046875,8.100738192732617
+world_model,16,1.4683058261871338,1.469836711883545,0.0040259781864860475,0.4262073825706135,0.4263343811035156,0.0011030904976787386,0.020717125555331017,0.0206034760922193,0.001819485709171538,0.020569573579864008,0.0206553414463996,0.0008590819101852027,0.1401055756617676,0.1353824883699417,0.0909198485313344,0.1620009683749892,0.1598714590072631,0.048894380025050574,144.14951116388494,143.28109741210938,7.3891086071698835
+world_model,17,1.4679158600893887,1.4688453674316406,0.0039857246144438755,0.4154709442095323,0.4156572818756103,0.0008925925535443907,0.021378530020063527,0.0213962737470865,0.001676572238549098,0.021537361497228746,0.02178505808115,0.0008468082337345843,0.15315092151815238,0.1482845842838287,0.08977501588268319,0.17738372087478635,0.1760808080434799,0.049186820891110816,143.03552107377485,143.22610473632812,7.195480436069859
+world_model,18,1.4636064009232954,1.467790603637695,0.009060840501285566,0.4054564129222523,0.4053222239017486,0.0011746381228250056,0.02188241566446689,0.0220874100923538,0.0016042578526410949,0.02224655398590992,0.0224944148212671,0.000872030717406511,0.16699790107932955,0.1624549925327301,0.08850828702245617,0.19314886900511646,0.1926280558109283,0.04946462810815876,141.7437411221591,141.19619750976562,6.616271626175893
+world_model,19,1.455238233913075,1.4681460857391355,0.0436510409825785,0.39611437103965064,0.3965134024620056,0.0008338883105889081,0.023427708074450448,0.0232527274638414,0.00163717064298724,0.023448466238650354,0.0237502492964267,0.0008398103140509136,0.18104473433711307,0.1762234568595886,0.08730788998658849,0.20954290032386777,0.2098750621080398,0.049758833221683146,140.8643937544389,139.97669982910156,6.160755021529545
+world_model,20,1.468782663345337,1.4702510833740234,0.005670644681389768,0.3869872201572765,0.3869568407535553,0.0010571064902677326,0.02425133894112972,0.02450642734766,0.0015303652421015958,0.024300455539063927,0.0244884025305509,0.0008867102792210733,0.19608496332710434,0.1916147023439407,0.08609799710254873,0.22648370807821097,0.2256379127502441,0.05009802607176699,140.0468486439098,138.7318572998047,5.8753852259971895
+world_model,21,1.4644020470705899,1.4684264659881592,0.01066176297492752,0.3791032135486603,0.3791325390338897,0.0009433582004157945,0.02539704011922528,0.0257000997662544,0.0016270135547439561,0.025749923322688387,0.0260105468332767,0.0008533973331368731,0.21189808980985117,0.2079691290855407,0.08469007303462316,0.2441297430883754,0.240778237581253,0.05037015641996531,139.3044794256037,138.02847290039062,5.5326629930242275
+world_model,22,1.461334683678367,1.4696552753448486,0.02302452352221058,0.370985209941864,0.3711584806442261,0.0014443361677846643,0.026766397566957863,0.0269958563148975,0.0016607692852825151,0.026968680660832956,0.0270332023501396,0.0008647437887070245,0.22820508886467322,0.2244424521923065,0.08320518329297112,0.26242396641861304,0.2565734088420868,0.05060509446514394,138.72936734286222,137.496337890625,5.132451633032861
+world_model,23,1.4446750337427312,1.4686152935028076,0.056259382419305566,0.36329191923141474,0.3630924522876739,0.0011915521634358126,0.028332868929613654,0.0285168830305337,0.0018320830224422342,0.028161855753172482,0.028290007263422,0.0009003314561876718,0.2448260513218966,0.2405172735452652,0.08171759805970372,0.2813629494471983,0.2730184197425842,0.05075338784327375,138.08518427068537,136.60797119140625,4.9349721352601215
+world_model,24,1.466230110688643,1.469010829925537,0.007285939344463773,0.3558488163081082,0.3559907376766205,0.0007745901520853519,0.02939577078954733,0.029836818575859,0.0018866230109178908,0.02951243011788884,0.0297594536095857,0.0009872368400971573,0.2629585672508586,0.2589176893234253,0.0803469479190814,0.3010982952334664,0.2910718321800232,0.050878586548391225,137.27650035511363,135.9857635498047,4.506471059784849
+world_model,25,1.4692075469277122,1.469456911087036,0.0027019828717012554,0.34885356101122766,0.3487524092197418,0.000793048342409851,0.030780516226183243,0.0310555100440979,0.001962576051029016,0.03095861863006241,0.0312995575368404,0.001026925046709204,0.2820501855828545,0.2783641815185547,0.07892049205094269,0.3216627402739091,0.3133932650089264,0.05100963034371276,136.63833479447797,135.25645446777344,4.2794082005690415
+world_model,26,1.4695656299591064,1.4703044891357422,0.002807688969282064,0.3418133150447499,0.3419051766395569,0.000663685055175396,0.032375999641689336,0.032511930912733,0.001968917488683674,0.03217066299508915,0.0325932428240776,0.0011124607390631982,0.30178275975314056,0.2981634736061096,0.07756046119040512,0.34296415339816694,0.3364678025245666,0.05105021280467984,136.08753551136363,134.88333129882812,3.9580865946858665
+world_model,27,1.466918858614835,1.4680020809173584,0.004439184246641668,0.3351816182786768,0.3350742161273956,0.0010690314138525696,0.034008620306849445,0.0343396440148353,0.0020573327960232146,0.033730834722518865,0.0340642482042312,0.0011571588361080083,0.32274481247771863,0.3197938203811645,0.07623022923330382,0.36515500057827344,0.3604071736335754,0.0510870348177181,135.60007268732244,134.2196044921875,3.753895564380846
+world_model,28,1.4683418273925781,1.4689431190490725,0.0029149424183031033,0.3292718854817477,0.3294259309768677,0.0007730130965883605,0.03559493916955858,0.0358923934400081,0.0021421180194335677,0.03523579815571954,0.0356733910739421,0.0012601516805333716,0.344605267047882,0.3428462147712707,0.07471696669486305,0.3882711448452689,0.3852652311325073,0.05096137085840122,135.177470814098,133.90113830566406,3.56739331406319
+world_model,29,1.4692777070132168,1.4690463542938232,0.001187872526910998,0.3227420801466161,0.3230057954788208,0.001070459439323339,0.037231924181634686,0.0375884436070919,0.0022783405771256955,0.0368433371186256,0.0374899692833423,0.0014017675875049184,0.3680680150335485,0.3676814436912536,0.07300795454030334,0.4123647592284463,0.4109280705451965,0.05070404195725256,134.5359649658203,133.32884216308594,3.3464185080870608
+world_model,30,1.4681182991374622,1.468722581863403,0.00259506658639151,0.3163115273822437,0.316324234008789,0.0008402644246002326,0.03909571231766174,0.0396018363535404,0.002318673557470767,0.03887488015673371,0.0395915359258651,0.0015070013733400952,0.3925800160928206,0.3935781717300415,0.07134856883751282,0.4376379739154469,0.4374896287918091,0.050258370700779396,134.09119762073863,132.93063354492188,3.107749833274907
+world_model,31,1.467529903758656,1.4698154926300049,0.004923710539639992,0.3114848163994876,0.3117312490940094,0.0007017182585621417,0.04102725637230003,0.0413359366357326,0.002523765263155732,0.040638839656656406,0.0412527583539485,0.00160690799148366,0.4189209152351726,0.417654275894165,0.06928672071710922,0.4639845273711465,0.4650555253028869,0.049640179448834795,127.844538601962,126.72604370117188,3.0088400327302525
+world_model,32,1.4503356326710095,1.4693527221679688,0.05764846954614597,0.3037180033597079,0.3036509156227112,0.0008390467655600075,0.04332004047252911,0.0436531417071819,0.0024305822702222354,0.041927914050492315,0.0425391420722007,0.0017415286135238277,0.44624681906266644,0.4415348768234253,0.06722988645750125,0.4913081201640042,0.493820309638977,0.048959126052730874,121.62532945112748,120.6963348388672,3.0135076768621607
+world_model,33,1.4609948288310657,1.469304084777832,0.02352221934185366,0.2956101840192621,0.295518010854721,0.0008179755114781201,0.04567472670565948,0.0471354387700557,0.003108594069972314,0.044044459407979734,0.0445928685367107,0.0018964095891150392,0.4764591401273554,0.4747952818870544,0.06434791967974723,0.5199022834951227,0.5237415432929993,0.04809487534011724,115.45348843661222,114.8263702392578,2.8654532731499383
+world_model,34,1.4661979458548806,1.468505859375,0.006777600232731323,0.28719224983995606,0.287119448184967,0.001012734486069327,0.047942356291142324,0.0493283085525035,0.003235846493284003,0.04606559771028427,0.0464515797793865,0.0019911458120029164,0.5079170573841442,0.5105005502700806,0.06134848933698819,0.5496855296871879,0.5546839237213135,0.04701157013568311,109.22610126842152,108.7387466430664,2.7526646745642505
+world_model,35,1.4654092138463801,1.4693818092346191,0.012530698601289168,0.27888039567253803,0.2789421379566192,0.0006728318857016968,0.050463817675005254,0.0518810600042343,0.003419995965195298,0.04802538894794198,0.0485400147736072,0.002179182625579962,0.5408043644644998,0.5478823184967041,0.058017562216878274,0.5805540626699274,0.5865806937217712,0.045679922235615726,103.01394306529652,102.85132598876952,2.652602642824873
+world_model,36,1.4671031128276477,1.4683687686920166,0.004608793628741282,0.27017998966303736,0.2702099084854126,0.0009087867604623105,0.053130560639229635,0.0543903075158596,0.003740563313714995,0.05109185217456381,0.0513737499713897,0.002458960536174388,0.5753087563948198,0.5833597779273987,0.054336592833437544,0.6129930290308866,0.6195119619369507,0.04387614533470697,96.8900541825728,96.91385650634766,2.5905874174665464
+world_model,37,1.468663519079035,1.4689886569976809,0.003353870088695131,0.26107886975461786,0.2611694931983948,0.0008109635240330389,0.055942094800147106,0.0564972162246704,0.004285833768827432,0.052179370414126944,0.0525184869766235,0.0025372012194642404,0.6120259924368425,0.6207063794136047,0.04998567422272302,0.6459511464292352,0.6535569429397583,0.04210749318734442,90.59552834250711,90.57064056396484,2.5087993264323707
+world_model,38,1.4685575311834163,1.4676339626312256,0.0029153948015766934,0.251905620098114,0.2516639530658722,0.000840152743565899,0.05845277959650209,0.0593115575611591,0.004573886376073989,0.05563866922801187,0.0555773228406906,0.0029281012174509975,0.6500857418233698,0.6595547199249268,0.045218634086322304,0.6803404905579307,0.6881426572799683,0.03963201004585964,84.4117001620206,84.40792083740234,2.4364240365385266
+world_model,39,1.466626297343861,1.4688293933868408,0.006235687915814369,0.24205284633419727,0.2419712096452713,0.0012442462959395325,0.06130189685658971,0.0624066442251205,0.004907480072507596,0.057563924653963545,0.0572689920663833,0.003131500885403345,0.68940654667941,0.7002094388008118,0.04014362993952567,0.7153850468722257,0.7237101793289185,0.036967243662913825,78.31840931285511,78.59146881103516,2.440420913268755
+world_model,40,1.4666812853379683,1.468860149383545,0.005317464354795765,0.2321554205634377,0.2322941720485687,0.0009222710606795902,0.064174396070567,0.0648738369345665,0.0052178271590239645,0.05857762728225096,0.0581852979958057,0.0032625578910711064,0.729470735246485,0.7402344942092896,0.03487963789572558,0.7501768307252363,0.7591898441314697,0.03421601789699089,72.19643055308948,72.60140991210938,2.4038235099128484
+world_model,41,1.4600811221382834,1.4689767360687256,0.027510905507638773,0.22236629507758396,0.2221569418907165,0.0011612707225271718,0.06885546716776757,0.0690470784902572,0.0054119027121710085,0.06224348091266366,0.0614409334957599,0.003788903358596341,0.7702723687345331,0.7792396545410156,0.029232089521372646,0.7856205864386125,0.7941232323646545,0.030718510559182035,66.1201882795854,66.51557922363281,2.3604195706790425
+world_model,42,1.420821493322199,1.4660840034484863,0.12309930931101802,0.21225083009763193,0.2121341824531555,0.0006794226917356145,0.07020466605370691,0.0703973770141601,0.0054573198403057,0.06412990458986972,0.0630325004458427,0.0040162280882026,0.810104949907823,0.8171495199203491,0.024048661879723997,0.8204707990993153,0.8286294937133789,0.02706637842966298,59.98217565363104,60.28294372558594,2.3424394118458687
+world_model,43,1.3999147848649458,1.469254970550537,0.21781755323622595,0.201913679187948,0.2019428163766861,0.0010410762372436981,0.0725201530890031,0.0733591392636299,0.0054560703243076375,0.06670289317315271,0.0654317140579223,0.004435684288902853,0.848534806208177,0.853638768196106,0.01919457673872723,0.8544365763664246,0.8620296716690063,0.023104866360112424,53.92282971468839,54.19173431396485,2.3018759645793496
+world_model,44,1.3944376165216619,1.4681918621063232,0.23179871564208238,0.1918082237243652,0.1917132884263992,0.0011266145758682539,0.07411486790938808,0.0744659677147865,0.0053340624463093615,0.06928501210429448,0.0674033015966415,0.005014592152386143,0.8842862573536959,0.8876720070838928,0.014797954371457913,0.8867509202523665,0.8930128812789917,0.01875954773917384,47.76955067027699,47.98651123046875,2.1773311004029763
+world_model,45,1.3962215943769976,1.4676592350006104,0.2336440161230607,0.18133696642788968,0.1811342984437942,0.0011126272096607466,0.07610171288251871,0.0750593394041061,0.0052231679024188175,0.07038058814677318,0.0686565786600112,0.0052372791660646265,0.9166603738611395,0.9177016615867616,0.010872336039513642,0.9161591854962435,0.921278953552246,0.014550500550090738,41.6348710493608,41.83378601074219,2.0061604610575925
+world_model,46,1.3959437933835117,1.4677042961120603,0.2334022765221959,0.1709649183533408,0.170794740319252,0.0009963828332423168,0.07662165910005567,0.0747245773673057,0.005026059803525407,0.0733391405506567,0.0711352825164794,0.0056793837645177404,0.9443203156644647,0.9427329897880554,0.007506113742280097,0.9426048723134127,0.946301519870758,0.010402146317958022,35.394454609264024,35.55489730834961,1.8793098762078355
+world_model,47,1.3919809081337668,1.4672982692718506,0.22987843257648233,0.16032402759248557,0.1601537615060806,0.001161449976876946,0.07885337891903786,0.0767847225069999,0.00498972098124497,0.07695143127983263,0.0743001103401184,0.006415993376916115,0.9672025279565291,0.9668025970458984,0.0046594976464106605,0.9653134400194342,0.9674853682518004,0.006501365069714958,29.084233544089578,29.25230026245117,1.7000212458647328
+world_model,48,1.3910502953962847,1.4627628326416016,0.2319554021951319,0.14953668144616208,0.1492391377687454,0.0010335816728866416,0.08014944602142676,0.0796511098742485,0.005092453726361779,0.07987065748734902,0.0776736214756965,0.0069555070170309605,0.9843516566536643,0.9842669367790222,0.002386530317992733,0.9830300320278514,0.9840595722198486,0.0032879741632763707,22.524729295210406,22.64668846130371,1.3865496963055841
+world_model,49,1.394185109571977,1.4651310443878174,0.23014187168674302,0.13735197755423456,0.1375972926616668,0.001190005738145348,0.0817304903810674,0.0838558226823806,0.005614900667317128,0.0799507004293528,0.0789465606212616,0.00683753993571891,0.9953602010553534,0.9953088164329528,0.0007703843441456976,0.99452528628436,0.9947948455810548,0.0011420244573810449,15.268142700195312,15.346698760986328,0.95061216057606
+world_model,50,1.3988349654457786,1.4692156314849854,0.2322583024738879,0.11424645578319371,0.1143975555896759,0.0020879029661737196,0.0990189205516468,0.1001605242490768,0.007705735583356467,0.10550027137452901,0.1046335324645042,0.010376522856520558,0.9999999891627919,1.0,6.429670929923101e-08,0.9999999729069796,1.0,7.233379796249756e-08,0.0,0.0,0.0

unitree_g1_pack_camera/case3/output/inference/round_summary.csv

@@ -0,0 +1,12 @@
+sample_id,scene,round_id,policy_pass_total_time_s,world_model_pass_total_time_s,round_total_time_s,latent_init_dist_to_prev_round,action_drift_vs_prev_round,psnr_full50
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,0,75.38506054878235,74.50816583633423,151.84783244132996,,,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,1,74.49218130111694,74.62240505218506,152.25997352600098,224.97146606445312,0.2796536087989807,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,2,74.69758439064026,74.56192207336426,151.35133409500122,228.21055603027344,0.15541577339172363,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,3,74.7608425617218,74.91416335105896,152.6145670413971,236.63555908203125,0.25457870960235596,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,4,74.69965195655823,74.96893620491028,151.8847451210022,231.79010009765625,0.918834924697876,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,5,74.81953477859497,74.73957371711731,152.68081712722778,198.5364227294922,0.08309799432754517,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,6,74.90933179855347,74.88512587547302,152.9385278224945,231.7943572998047,0.11995601654052734,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,7,74.84911894798279,74.95093822479248,152.97526359558105,236.86546325683594,0.39963626861572266,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,8,74.88821697235107,67.77415251731873,145.36549997329712,240.78086853027344,0.4154035449028015,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,9,35.5922589302063,64.26668500900269,101.50574612617493,198.4535675048828,0.06449002027511597,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,10,74.76077079772949,74.81333827972412,152.70444345474243,124.39263153076172,0.005869686603546143,

unitree_g1_pack_camera/case3/output/inference/sample_summary.csv

@@ -0,0 +1,3 @@
+sample_id,scene,pass_type,pass_total_time_s,action_first_stable_step,state_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_state,oracle_budget_latent,latent_init_dist_to_prev_round,action_drift_vs_prev_round,round_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,psnr_full50
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,71.25950481674887,,,,44.72727272727273,46.63636363636363,46.63636363636363,46.72727272727273,,215.24309921264648,0.2696936547756195,147.10261366584083,71.25950481674887,73.18230964920737,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,73.18230964920737,,,,45.0,46.72727272727273,46.72727272727273,46.72727272727273,,215.24309921264648,0.2696936547756195,147.10261366584083,71.25950481674887,73.18230964920737,

unitree_g1_pack_camera/case3/output/sparse_8/inference/round_summary.csv

@@ -0,0 +1,12 @@
+sample_id,scene,round_id,policy_pass_total_time_s,world_model_pass_total_time_s,round_total_time_s,latent_init_dist_to_prev_round,action_drift_vs_prev_round,psnr_full50
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,0,34.73047733306885,34.12665319442749,143.00190567970276,,,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,1,34.77882504463196,35.024171590805054,142.87922620773315,238.14430236816406,0.34216785430908203,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,2,34.65710210800171,34.31918025016785,141.75025510787964,244.6288299560547,0.10849124193191528,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,3,34.64808392524719,34.763160705566406,142.0598931312561,250.288330078125,0.4585834741592407,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,4,34.43521761894226,34.34122323989868,141.00554966926575,224.81471252441406,0.6429577469825745,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,5,34.357038497924805,34.85403919219971,141.1495542526245,195.2209014892578,0.1021263599395752,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,6,35.0233428478241,34.67922353744507,144.0218186378479,226.24966430664062,0.08848029375076294,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,7,34.45149564743042,34.6253707408905,143.35860419273376,245.75660705566406,0.17197155952453613,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,8,34.7300283908844,34.399852991104126,143.3491985797882,238.90292358398438,0.6555709540843964,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,9,34.11410880088806,34.445759534835815,142.07709193229675,226.96424865722656,0.0632086992263794,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,10,34.794456481933594,34.530436992645264,141.20812225341797,223.38699340820312,0.012062489986419678,

unitree_g1_pack_camera/case3/output/sparse_8/inference/sample_summary.csv

@@ -0,0 +1,3 @@
+sample_id,scene,pass_type,pass_total_time_s,action_first_stable_step,state_first_stable_step,latent_first_stable_step,action_vs_full50_90pct_step,action_vs_full50_95pct_step,oracle_budget_action,oracle_budget_state,oracle_budget_latent,latent_init_dist_to_prev_round,action_drift_vs_prev_round,round_total_time_s,policy_pass_total_time_s,world_model_pass_total_time_s,psnr_full50
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,policy,34.610925154252485,,,,45.54545454545455,47.45454545454545,46.72727272727273,46.72727272727273,,231.43575134277344,0.26456206738948823,142.35101996768606,34.610925154252485,34.555370179089635,
+unitree_g1_pack_camera-vid100-fs6,unitree_g1_pack_camera,world_model,34.555370179089635,,,,45.90909090909091,48.0,46.90909090909091,46.72727272727273,,231.43575134277344,0.26456206738948823,142.35101996768606,34.610925154252485,34.555370179089635,

unitree_g1_pack_camera/case3/run_world_model_backbone_profile.sh

@@ -0,0 +1,11 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+
+repo_root="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
+CUDA_VISIBLE_DEVICES="${CUDA_VISIBLE_DEVICES:-1}" \
+bash "${repo_root}/scripts/run_world_model_backbone_profile_case.sh" \
+    "unitree_g1_pack_camera/case3" \
+    "unitree_g1_pack_camera" \
+    "6" \
+    "11"

unitree_g1_pack_camera/case3/run_world_model_interaction.sh

@@ -0,0 +1,29 @@
+res_dir="unitree_g1_pack_camera/case3"
+dataset="unitree_g1_pack_camera"
+
+{
+    time CUDA_VISIBLE_DEVICES=0 "${PYTHON_BIN:-python}" scripts/evaluation/world_model_interaction.py \
+        --seed 123 \
+        --ckpt_path ckpts/unifolm_wma_dual.ckpt \
+        --config configs/inference/world_model_interaction.yaml \
+        --savedir "${res_dir}/output/sparse_8" \
+        --bs 1 --height 320 --width 512 \
+        --unconditional_guidance_scale 1.0 \
+        --ddim_steps 50 \
+        --ddim_eta 1.0 \
+        --prompt_dir "unitree_g1_pack_camera/case3/world_model_interaction_prompts" \
+        --dataset ${dataset} \
+        --video_length 16 \
+        --frame_stride 6 \
+        --n_action_steps 16 \
+        --exe_steps 16 \
+        --n_iter 11 \
+        --timestep_spacing 'uniform_trailing' \
+        --guidance_rescale 0.7 \
+        --perframe_ae \
+        --analysis_log_metrics \
+        --analysis_reference_steps 50 \
+        --head_schedule_steps 0 7 14 21 28 35 42 49 \
+        --head_skip_mode reuse_prediction \
+        --head_log_steps 40 43 46 47 48 49
+} 2>&1 | tee "${res_dir}/output.log"