Two predictor calls fix world-model counterfactuals: R² jumps from 38–46% to 86–96%

TheTuringPost · x · 2026-11-21

A world-model research thread: standard JEPA-style models trained on factual data only return the average outcome when asked 'what if I'd acted differently', discarding information about hidden disturbances. The fix exploits that action–disturbance interaction is second-order in time: estimate counterfactual outcome as observed + prediction(other action) − prediction(taken action).

Key results:

The authors prove an O(h²) error bound and that no method using only unpaired factual data can achieve a better rate. They also show where it breaks: contacts, folded coordinates, and faults that scale the command.

Related event: LeCun and Colleagues Release Counterfactual World Models Paper(12 posts)→

Original post →

More from Research

Research channel →