← back to paper

Review history

arxiv: 2605.11854 · 2 revisions

Self-Distilled Trajectory-Aware Boltzmann Modeling: Bridging the Training-Inference Discrepancy in Diffusion Language Models

1. 2026-05-20 UNVERDICTED LOW v0.9.0 novelty 7.0
   45578 ms 5869 in 1387 out 2026-05-20T22:55:59.330682+00:00
2. 2026-05-13 UNVERDICTED LOW v0.9.0 novelty 6.0
   36998 ms 5638 in 1320 out 2026-05-13T05:48:46.216040+00:00