2026-10-27 · ArXiv
arXiv:2610.35425v1 Announce Type: new
Abstract: Predictive uncertainty in stochastic language-model inference can be characterized by repeatedly sampling an input and measuring the distribution of outputs. This requires multiple complete generations. Masked diffusion language models expose another source of information: the evolution of the output during denoising. We ask whether convergence within diffusion trajectories is associated with uncertainty across their final outputs, or endpoints. We study four masked diffusion language models on four reasoning benchmarks, collecting 20 generations per question and 32,000 denoising trajectories in total. Within each trajectory, we measure within-run concentration as the area under the curve describing stable resolution of the eventual output. Across repeated trajectories, we measure endpoint-distribution uncertainty. Stronger within-run concentration is associated with lower endpoint entropy after accounting for model and dataset ($\beta=-0.222$, $p
原文链接：https://arxiv.org/abs/2610.35425
查看更多资讯请见 AI 行业资讯。