TheMachine Press

A daily newspaper for the age of artificial intelligence.

Morning editionPermanent story

research

The World Model Learned Energy and Let It Drift

A frozen video world model encoded an energy-like invariant that its own imagined rollouts failed to preserve.

Published Updated Story ID: mp-2026-08-25-005
Read the complete editionStory JSON

Summary

A frozen video world model encoded an energy-like invariant that its own imagined rollouts failed to preserve.

A label-free search recovered the same scalar across independently trained conservative pendulum models and found no comparable quantity in damped controls. Projecting latent state back toward the initial level set reduced rollout error in all three conservative models; matched random constraints usually made it worse.

Why it matters

A frozen video world model encoded an energy-like invariant that its own imagined rollouts failed to preserve.

Limits and context

No additional limitation was separately recorded.

Key claims

  1. A frozen video world model encoded an energy-like invariant that its own imagined rollouts failed to preserve.

    Evidence: source-2026-08-25-005

Sources

  1. arXiv preprint 2608.23526arXiv · primary research

Corrections

No corrections have been recorded for this story.