Reproduce “DiMuRe: Diagnosis-Guided Multilevel Recovery for Robotic Task Execution” 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
Renkai Yi, Yuwei Wu, Haoyu Zhang, Penghao Dai, Zhi Gao, Jin Chen, Che Sun
DiMuRe chooses among adapting reusable program actions, trying eligible recovery skills, and generating corrective code with a multimodal language model after the original program exits. It verifies local recovery before continuing. On 801 scheduled disturbed cases in simulation, local recovery achieved 77.65% task success versus 58.68% for the program under the physical-step allowance; this subset was not held out.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.