Reproduce “Error-Feedback Model for Output Correction in Bilateral Control-Based Imitation Learning” 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
Hiroshi Sato, Masashi Konosu, Sho Sakaino, Toshiaki Tsuji
The method adds feedback to a two-layer imitation model by comparing the upper layer’s expected next state with the lower layer’s prediction and feeding the difference back during operation. The lower layer, trained only on “A,” follows supplied plans for unfamiliar characters. For “4,” MLP IoU was 0.204 ± 0.065 with feedback versus 0.127 ± 0.034 without; “B” remained poor.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.