Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2603.28955 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912989987209216 |
|---|---|
| author | Han, Yuci Yilmaz, Alper |
| author_facet | Han, Yuci Yilmaz, Alper |
| contents | This paper presents the World-Action Model (WAM), an action-regularized world model that jointly reasons over future visual observations and the actions that drive state transitions. Unlike conventional world models trained solely via image prediction, WAM incorporates an inverse dynamics objective into DreamerV2 that predicts actions from latent state transitions, encouraging the learned representations to capture action-relevant structure critical for downstream control. We evaluate WAM on enhancing policy learning across eight manipulation tasks from the CALVIN benchmark. We first pretrain a diffusion policy via behavioral cloning on world model latents, then refine it with model-based PPO inside the frozen world model. Without modifying the policy architecture or training procedure, WAM improves average behavioral cloning success from 59.4% to 71.2% over DreamerV2 and DiWA baselines. After PPO fine-tuning, WAM achieves 92.8% average success versus 79.8% for the baseline, with two tasks reaching 100%, using 8.7x fewer training steps. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_28955 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Enhancing Policy Learning with World-Action Model Han, Yuci Yilmaz, Alper Artificial Intelligence This paper presents the World-Action Model (WAM), an action-regularized world model that jointly reasons over future visual observations and the actions that drive state transitions. Unlike conventional world models trained solely via image prediction, WAM incorporates an inverse dynamics objective into DreamerV2 that predicts actions from latent state transitions, encouraging the learned representations to capture action-relevant structure critical for downstream control. We evaluate WAM on enhancing policy learning across eight manipulation tasks from the CALVIN benchmark. We first pretrain a diffusion policy via behavioral cloning on world model latents, then refine it with model-based PPO inside the frozen world model. Without modifying the policy architecture or training procedure, WAM improves average behavioral cloning success from 59.4% to 71.2% over DreamerV2 and DiWA baselines. After PPO fine-tuning, WAM achieves 92.8% average success versus 79.8% for the baseline, with two tasks reaching 100%, using 8.7x fewer training steps. |
| title | Enhancing Policy Learning with World-Action Model |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2603.28955 |