Data for Arnold: A multi-task, multi-embodiment muscle transformer policy
Authors/Creators
Description
Supplementary Data — Trained Model Checkpoints and Benchmark Results
====================================================================
This archive set accompanies an anonymous manuscript submission. It contains
the final trained model weights and the corresponding benchmark evaluation
results for every model reported in the paper.
Contents
--------
1. arnold-final-checkpoints.tar.gz.
Trained model weights. Unpacks to: <Arnold-root>/data/final_checkpoints/
~2.8 GB. One subdirectory per model (see "Models" below), each
containing per-seed subdirectories (seed_0, seed_1, ...).
A typical seed directory contains:
- rl_model_<steps>_steps.zip Stable-Baselines3 policy checkpoint
- rl_model_vecnormalize_<steps>_steps.pkl VecNormalize statistics
- model_config.json / args.json training + model configuration
- <Task>_config.json per-task environment configs
- vocabulary.json sensorimotor vocabulary (where used)
- main_bc_ppo_multi_task.py training entry point (snapshot)
- PPO_0/ training logs / tensorboard events
2. arnold-benchmark-results.tar.gz
Benchmark evaluation results. Unpacks to: <Arnold-root>/data/final_benchmarks/
~55.5 MB. Same model/seed layout as above; each seed directory holds:
- seed_<n>_<checkpoint>_results.json per-task benchmark scores
- *.log evaluation run logs
Also contains the expert evaluation results.
3. arnold-benchmark-results-extra.tar.gz
Benchmark evaluation results. Unpacks to: <Arnold-root>/data/final_benchmarks_extra/
~11 MB. Contains files used by plotting or other analysis in the paper.
4. expert-policies.tar.gz
Expert checkpoints. Unpacks to: <Arnold-root>/data/expert_policies/
~605 MB.
5. analysis.tar.gz
Files used for EMG, smoothness, PCA and other analysis. Unpacks to <Arnold-root>/data/analysis/
~318 MB.
6. kinesis.tar.gz
Model files and data for the Walk-to-point (Kinesis) task. Unpacks to: <Arnold-root>/data/kinesis/
~29 MB.
File description for final_checkpoints and final_benchmarks
|
Model |
Description |
|---|---|
|
|
Proposed method (multi-task compositional policy) |
|
|
On-policy behavior cloning baseline |
|
|
OBC trained on the 10-task subset |
|
|
OBC capacity variants (medium / small / extra-small) |
|
|
OBC single-task specialists |
|
|
OBC, task-specific sensorimotor vocabulary variant |
|
|
OBC with PPO fine-tuning |
|
|
OBC ablation without observation normalization |
|
|
Behavior-cloning baseline |
|
|
Transformer-based PPO (no sensorimotor vocabulary) |
|
|
Transformer-based PPO with sensorimotor vocabulary |
|
|
Multi-task PPO |
|
|
Multi-task SAC |
Files
ZENODO_README.txt
Files
(3.8 GB)
| Name | Size | |
|---|---|---|
|
md5:75a8b1b71cc0a3b1d587fde2bcd01b3b
|
318.2 MB | Download |
|
md5:2e6d489d1c4a2905647b118806fe4202
|
2.9 MB | Download |
|
md5:22465af9735dc2ee3d1d18c644a2afc8
|
54.6 MB | Download |
|
md5:6c5fa8da01f47ad0d5a17ce21a8b5bdb
|
2.8 GB | Download |
|
md5:94ecedb633d251becb70249b88fffb4f
|
605.1 MB | Download |
|
md5:2479581598dabeb78894213f0e721c61
|
13.1 MB | Download |
|
md5:db00505dfe8debee8053a84769ed491b
|
2.0 kB | Preview Download |
Additional details
Related works
- Is supplement to
- Preprint: arXiv:2508.18066 (arXiv)
Funding
- Swiss National Science Foundation
- 310030_212516
- Simons Foundation
- SFI-AN-NC-SCN-00007276-14