Reproduce “CoBoVLA: Decoupling Language from Action with Discrete Codes for Robust and Interpretable Robot Manipulation” 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

Mehul Arora, Dhruv Agarwal, Ambuj Agrawal

CoBoVLA separates instruction understanding from robot control using discrete step codes, motion-segment codes, and an actor that receives no language. Its text encoder maps paraphrases to shared task codes, while a visual planner selects motion codes from images and robot state. In simulation, it achieved 81.1% success across 4,092 LIBERO-Para paraphrases; code agreement was not universal.

Summary by alphaXiv.

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