This paper introduces LingBot-World 2.0, an interactive world model that can generate unbounded, stable video streams at 720p and 60 fps. The system uses a causal pretraining paradigm to resist error accumulation over long horizons, verified by over an hour of continuous generation with no quality decay. It supports a wide range of character actions (combat, archery, spell-casting, shooting) and environmental modifications (e.g., weather changes). The model employs an agentic harness with a pilot agent for character behavior and a director agent for synthesizing new scene elements. A 14B base model and a distilled 1.3B variant are provided, with the smaller model deployable on a single GPU.
World simulator achieves infinite interaction and agentic control
LingBot-World 2.0 introduces causal pretraining for unbounded horizons, real-time 60fps rendering, diverse actions, and a dual-agent harness for planning and environment synthesis.
Independent
Zelin Gao · Qiuyu Wang · Jiapeng Zhu · Jingye Chen · Zichen Liu · Qingyan Bai · +14 more
Research Digest··2 min read
In one line
LingBot-World 2.0 generates infinite, real-time 720p 60fps interactive worlds with diverse actions and environmental changes.
Before you read this
Three ideas this paper builds onlatent world modelA learned internal, compressed representation of environment dynamics, enabling future-aware planning, generation, and reasoning directly in latent space.44 papers →multi-agent architectureA system where multiple AI agents collaborate, each handling specialized roles, communicating and coordinating to complete complex tasks.99 papers →agent-driven simulationAutonomous agents automatically generate, control, or extend simulations to improve accuracy, interactivity, or performance.32 papers →
Read it as
The authors present LingBot-World 2.0, an upgrade to their open-world simulator.
The authors developed LingBot-World 2.0, an enhanced world simulator built on a 14B parameter base model paired with a lightweight 1.3B variant for single-GPU deployment.
The Daily Edition
The front page in your inbox each weekday morning: the lead, the picks, and why each one earned its place.