Reasoning on the Manifold: Bidirectional Consistency for Self-Verification in Diffusion Language Models 事件

Reasoning on the Manifold: Bidirectional Consistency for Self-Verification in Diffusion Language Models arXiv:2604.16565v3 Announce Type: replace-cross Abstract: While Diffusion Large Language Models (dLLMs) offer structural advantages for global planning, efficiently verifying that they arrive at correct answers via valid reasoning traces remains a critical challenge. In this work, we propose a geometric perspective: Reasoning on the Manifold. We hypothesize that valid generation trajectories