Reproduce “From Pretraining to Proficiency: Real-World Subtask RL for Long-Horizon Manipulation with Minimal Human Intervention” 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
Sichang Su, Benjamin Yang, Zhiyun Deng, Boyuan Liang, Yip Fun Yeung, Zelin Wang, Lingfeng Sun
PARTS improves long-horizon robot tasks by keeping a pretrained policy frozen and training bounded residual policies at selected bottlenecks. A selector activates each correction under specified conditions, while a verifier supplies local success or failure rewards. On a cable task, complete-task success was 95% (19/20), versus 50% (10/20) for the SFT base policy, across 20 episodes per method in three real-robot tasks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.