robotics
The Robot Improved When Its Reasoning Pointed to the Next Action
ARC trains foundation models on traces that explain why the immediate action fits and what effect it should produce.
Summary
ARC trains foundation models on traces that explain why the immediate action fits and what effect it should produce.
ARC adds action-grounded reasoning traces to existing robot foundation models without collecting new robot demonstrations. An automatic pipeline labels demonstrations with the causal structure of the next action, then adapts vision-language-action and world-action models to use those traces during control. The authors report state-of-the-art results on RoboLab-120 and MolmoSpaces, gains up to 50 percentage points on RoboLab-Reasoning-50 and an 82.2-point improvement for one adapted policy on real-robot tasks. Those figures are model- and benchmark-specific. The study's broader claim is that structured reasoning supervision can complement scale, not that verbal traces alone solve general robotic control.
Why it matters
ARC trains foundation models on traces that explain why the immediate action fits and what effect it should produce.
Limits and context
- The study's broader claim is that structured reasoning supervision can complement scale, not that verbal traces alone solve general robotic control.
Key claims
ARC trains foundation models on traces that explain why the immediate action fits and what effect it should produce.
Qualification: The study's broader claim is that structured reasoning supervision can complement scale, not that verbal traces alone solve general robotic control.
Evidence: source-2026-10-11-008
Sources
- arXiv preprint 2610.12386, original v1arXiv / submitting research authors · primary research
Corrections
No corrections have been recorded for this story.