Article

Feedback World Model Enables Precise Guidance of Diffusion Policy (NeurIPS 2026)

This paper proposes a feedback world model that uses the true states observed after robot execution to perform online closed-loop correction of predictions at inference time, without additional training data or parameter updates. The method can be interpreted as a latent-space observer and provides convergence guarantees under bounded, asymptotically convergent dynamics residuals and a stable feedback gain; it also introduces action-aware guidance to better translate corrected predictions into control.

Part of the event: Feedback World Model Proposed to Online-Correct Diffusion Policy Predictions

This page provides a summary and source link only.