Reproduce “RoboArchitect: Automating the Design of Robotic Agents with Persistent Memory” 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
Shengguang Wu, Suning Huang, Cherie Ho, Ria Doshi, Hao Zhu, Serena Yeung-Levy
RoboArchitect evolves control code around frozen robot policies by having a coding model inspect execution traces, diagnose failures, and revise a harness without updating policy weights. Across 10 policy–benchmark setups, mean reported scores rise from 9.7% to 43.3% on unseen scenes and from 12.7% to 40.2% on unseen tasks. Evaluations are in simulation; real-world transfer remains untested.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.