Benchmarks

WROP

Benchmark · 1 item since Sep 2026

Whether video world models keep track of objects that are hidden or out of view.

Items

  1. Sep 25

    WROP teaches world models object permanence

    Introduces WROP, a cognitive-science-inspired benchmark and 1.5M-sample corpus for object permanence, and trains a 16B video world model (PWM-WROP) that ranks first among continuation models in a blind pairwise Elo study.