Reproduce “SpikWM: Efficient End-to-End Spiking World Modeling for Visual Planning” 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

Shuhan Ye, Hongbin Yu, Chong Wang, Hui Lu, Jiaming He, Yi Yu, Jiangbo Qian, Xudong Jiang

SpikWM is a spiking world model for goal-conditioned planning from images. It preserves spatial cues, encodes positive and negative action responses, and carries membrane evidence across latent positions, then uses a planner to compare predicted endpoints with a goal. On Push-T, it reaches 88.67% success versus 53.33% for Spiking-LeWM, but remains below LeWM’s 90.4% mean.

Summary by alphaXiv.

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