Commit Locally, Exit Globally: Coordinating Adaptive Sampling and Early Exit in Diffusion Language Models
Diffusion language models predict tokens in parallel but still require many iterative forward passes. Adaptive sampling accelerates local token commitment, while recent global stopping methods seek to end generation early. Our trajectory analysis shows that short answers can stabilize early in decoding, whereas reasoning outputs often require more steps to settle. A fixed progress threshold therefore struggles to avoid both redundant computation on short-answer tasks and premature termination on reasoning tasks. We introduce C$^4$, a training-free framework that combines local decoding acceleration with adaptive global early exit. Confidence-Verified Early Exit (CVEE) determines when to e...
| Authors | Chia-Ming Lee, Shao-Kai Liu, Ming-Ching Chang et al. |
|---|---|
| Primary category | cs.CL |
| Published | 30 July 2026 |
| Last updated at source | 14 October 2026 |
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.