Trace-Focused Diffusion Policy for Multi-Modal Action Disambiguation in Long-Horizon Robotic Manipulation
Yuxuan Hu, Xiangyu Chen, Chuhao Zhou, Yuxi Liu, Gen Li, Jindou Jia, Jianfei Yang 1,† · 2026
Read Trace-Focused Diffusion Policy for Multi-Modal Action Disambiguation in Long-Horizon Robotic Manipulation by Yuxuan Hu, Xiangyu Chen, Chuhao Zhou, Yuxi Liu, Gen Li, Jindou Jia, Jianfei Yang 1,† (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-07388.txt.
Description
In long robot tasks, the same view can recur at other stages demanding other actions, confusing policies that look only at the present moment. Yuxuan Hu and six co-authors address this multi-modal action ambiguity with a Trace-Focused Diffusion Policy that conditions action generation on execution history.
Past motion is represented as an explicit trace projected into the visual observation space, supplying stage-aware context, while a trace-focused field emphasizes regions tied to that history. The policy thus knows not just what it sees but where it is in the task. Imitation learning teams get a simple, effective fix for ambiguity that plagues long-horizon manipulation.
Catalog record
| CREATOR | Yuxuan Hu, Xiangyu Chen, Chuhao Zhou, Yuxi Liu, Gen Li, Jindou Jia, Jianfei Yang 1,† |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 42 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-07388