PACE-dLLM: Elastic Block Decoding via Confidence Cliff Estimation for Diffusion Language Models
PACE-dLLM is proposed, which fits this parametric cliff in closed form at each step, sets the next horizon by its saturation point, and uses an independent confidence threshold for token commitment, and achieves the best average accuracy on both open-source dLLM backbones.
Xiao-Cheng Lu, Shu-Han Guo, Zi-Yue Ma et al.
· 1 citation