Reproduce “LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning” 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
Xuheng Zhang, Yiming Zhang, Jialin Gao, Hao Chen, Jiaming Liu, Ji Li, Runzhong Huang, Yinxi Wang and 9 more
LaST₀.₅ adds an interaction-focused latent representation between predicted scene dynamics and action generation, using motion and target-object affordance cues during training. Its adaptive executor uses predicted task progress to decide whether to continue an action chunk or replan. It reports 99.5% across four LIBERO’s suites and 88% across ten RLBench tasks, averages that do not imply success on every task.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.