Reproduce “PlanContext: Benchmarking In-Context Driving Decision-Making” 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
Ruikai Li, Naiyu Gao, Hao Shan, Yizhe Fan, Yanhao Wu, Han Jiang, Yilong Ren, Zhenning Li and 6 more
PlanContext evaluates whether end-to-end driving policies respond appropriately when navigation intentions change in a fixed scene. It compares predictions under original and alternative commands, scoring whether vehicles follow, wait, reject, or preserve a maneuver while meeting road and safety criteria. Across 23 models, mPDMS and NCS have a Spearman correlation of −0.542, showing differing rankings in offline evaluation.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.