RULER-Bench: Probing Rule-based Reasoning Abilities of Next-level Video Generation Models for Vision Foundation Intelligence Paper • 2512.02622 • Published 24 days ago • 9