Reproduce “Latent-WAM++: Shaping World and Action Representations for End-to-End Autonomous Driving” 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

Yichen Zhang, Yupeng Zheng, Linbo Wang, Pengxuan Yang, Zebin Xing, Junli Wang, Shiwei Li, Kangwei Liu and 5 more

Latent-WAM++ trains a camera-based driving planner by supervising scene features for future-state prediction, stabilizing their distributions, and aligning trajectory-producing features with visual features. Its planner generates trajectory proposals and refines them. On NAVSIM v2, it scores 90.1 EPDMS, versus 89.3 for Latent-WAM, while using 190 GPU-hours to train versus 1,600. These benchmark and simulator results do not demonstrate real-world deployment.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.