Reproduce “A computational framework for understanding embodied plans” 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
Aalap D. Shah, Basil Saeed, Joshua B. Tenenbaum, Tobias Gerstenberg, Hyowon Gweon, Ilker Yildirim
GenEA models how people infer goals from embodied actions and physical effects. It combines symbolic actions, kinematic trajectories, and simulated dynamics to assign probabilities to plans reaching specified outcomes. Across tested tasks, its predictions tracked human judgments, with navigation ratings correlated at r = .91, p < .001. This supports modeled physics in these tasks, not claims about the brain’s algorithm or robot performance.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.