SyncAI.news, a Varaisys broadcasting
ViStep-RL: Stepwise Reinforcement Learning for Long-Horizon Visual Agents
JC

Junyang Chen, Yao Qin, Xiao Li, Ningchao Wang, Yiguo Qiao, Wenbin Li, Xin Geng

· 1 min read

ResearcharXiv cs.CV

ViStep-RL: Stepwise Reinforcement Learning for Long-Horizon Visual Agents

arXiv:2611.24170v1 Announce Type: new Abstract: Long-horizon visual agents must do more than inspect images: they must turn observations into evidence that remains usable as a task unfolds. Existing active-perception methods teach agents where and how to look, while a final-answer reward gives little structure for deciding which observations satisfy dependent subgoals or should be retained for later use. We present ViStep-RL (Visual Stepwise Reinforcement Learning), which models execution as a dependency-linked graph of evidence-bearing subgoals. A verifier tracks which subgoals are supported by valid evidence, and changes in graph potential provide a net-progress learning signal; an active memory protocol preserves evidence across turns and links retrieval to downstream use. On ViInteract, ViStep-RL reaches 60.0% strict completion with Qwen2.5-VL-7B, compared with 10.0% for the base model and 46.5% after supervised fine-tuning. Controlled studies show that incremental task-state rewards outperform cumulative rewards by 3.2 points, while active evidence writing improves strict completion by 7.75 points over automatic writing. Under a common interaction framework, the task-trained ViStep-RL policy achieves 59.5% answer accuracy, 24.0 points above the strongest of six general-purpose reference models. Held-out episodes confirm the gain at both Qwen scales, and gains extend to three external visual reasoning benchmarks.

Original source

This story was published by arXiv cs.CV and written by Junyang Chen, Yao Qin, Xiao Li, Ningchao Wang, Yiguo Qiao, Wenbin Li, Xin Geng. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on arxiv.org

Similar News