# Rollback: Using World-Model Backtracking of Counterfactual Experience to Improve Robot Policies

- Source: HuggingFace Daily Papers
- Published: 2026-10-26T00:00:00.000Z
- RoboHot: https://robohot.news/en/items/bng8iztnueayt0co0ejk2zgg9
- Original: https://huggingface.co/papers/2610.36068

## Summary

Rollback proposes a policy improvement paradigm that turns an action-conditioned world model from a one-way simulator into a backtrackable counterfactual engine: it returns to key decision points, branches to try different actions, and compares the outcomes, thereby distilling counterfactual results into action-level supervision without learning a separate value model.

## Why it matters

The world model is used as a backtrackable counterfactual engine to provide fine-grained action-level supervision for policy improvement, and the improvement margins on three real-robot tasks are reported.