Programmable World Model: agent-programmed persistent state for interactive video

Alaya Lab

Research official 1 src. ~1 min

The framework decouples world-state evolution from visual generation: an agent compiles natural-language rules into executable state-transition programs, a lightweight engine maintains a persistent global state (including off-screen entities), and state-augmented 3D bounding boxes are deterministically compiled into conditioning for a pretrained video renderer. On the new CombatStateBench it hits 94% count accuracy and 98% state accuracy, far above existing interactive video world models.

Why it matters

81 upvotes on HuggingFace Daily Papers (Sep 10). Addresses the core unsolved problem of interactive video world models — persistent, rule-respecting world state — with a controllable, game-capable design and released code.

Importance: 3/5

Notable paper addressing persistent world state in interactive video generation

Sources