Reproduce “Human-Robot Collaboration via Deep Reinforcement Learning of Real-World Interactions” 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
Jonas Tjomsland, Ali Shafti, A. Aldo Faisal
The authors train a robot agent through real-world interaction with a person on a two-axis ball-and-tray task, with each controlling one axis. Training takes about 30 minutes, with 3,000 physical interaction steps and 60,000 offline gradient updates. In a ten-subject comparison, five showed no significant performance difference from human-human teams; five performed worse, so the finding does not establish equivalence.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.