Reproduce “Embedding Physics Priors in Robot Learning: A Survey” 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
Mattia Piccinini, Lucas Schulze, Alice Plebe, Matteo Saveriano, Thomas Beckers, Yuan Gao, Oleg Arenz, Baha Zarrouki and 5 more
The survey maps 232 physics-embedded robot-learning methods among 329 cited references, organizing them by whether physics enters inputs or representations, model architecture, or training loss. Physics-encoded architectures form 71% of methods by primary route, versus 16% physics-guided and 13% physics-informed. Only 4% combine routes, so the review does not establish that any approach performs best.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.