Read the Trace, Steer the Path: Trajectory-Aware Reinforcement Learning for Diffusion Language Models

ArXi:2606.04396v1 Announce Type: new Diffusion large language models (dLLMs) generate responses by iteratively unmasking and revising many positions in parallel. This process leaves a rich denoising trace depicting which tokens become confident, which remain unstable, and when commitments form. Existing dLLM reinforcement learning methods use this signal only weakly. Flat rollouts are cheap, but assign a single outcome reward to the whole trajectory. Tree rollouts provide finer, verifiable.