Reproduce “Relational Learning for Skill Preconditions” 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
Mohit Sharma, Oliver Kroemer
The authors learn pairwise relations from simulated interactions and compose them to predict skill preconditions in manipulation scenes. A separate predictor uses real-world task data; the relation model stays frozen. In food cutting, the GNN achieves weighted F1 0.960 versus 0.868 for ResNet-34, training with up to two distractors and testing with four; this measures precondition classification, not cutting success.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.