SOURCE-LINKED INTELLIGENCE

Verifiable Hidden Dynamics Play: Generating Agentic RL Environments from Solved Mechanisms

arXiv · AI, language, vision and robotics · article · Sep 23, 2026 · UTC

Language-model agents increasingly face long-horizon tasks with evolving state, interdependent decisions, and delayed outcomes. Scaling their training requires diverse agentic environments, dependable outcome signals, and low extension cost. Existing generation pipelines commonly construct an environment before defining its outcome rule or annotating its trajectories, leaving dynamics and evaluation to be aligned post hoc. VHD-Play reverses this dependency by sampling and solving a mathematical model before a corpus-grounded setter renders its decision process as stateful tools. The executable

Read original source ↗ Open in workspace

recordType

paper

region

Global