Reproduce “WAM is capable in BEHAVIOR-1K: the 2nd-place solution to the 2026 BEHAVIOR Challenge” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yelin He, Jiaquan Ye, Xinming Wei, Wenjie Xiao, Gengming Zhang, Guoting Wei, Xianbiao Qi, Rong Xiao
IF1001 uses Cosmos 3 visual history and estimated task progress to route household tasks to specialized policies. The system combines a world action model, task-group LoRA adapters, and execution controls. On the 2026 BEHAVIOR challenge’s 100-task private test, it ranked second, with a Q-score of 0.615 and full-task success of 41.2%; this result applies to the assembled system, not any single component.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.