Newest papers

ReAnchor: Reusing Rejected Verifier States for Diffusion Drafting

Sanghyun Lee, 남원진, Seungho Jang, Jaeho Lee, Jiho Park, Jaewoon Lee, Seungryong Kim · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.26531 · cited by 0 works (OpenAlex) · open-access PDF

Korean

Block diffusion drafters는 단일 forward pass에서 여러 draft 토큰을 생성함으로써 추측적 해독을 가속화합니다. 그러나 각 pass는 단일 target-generated anchor에서 단일 고정 길이 블록을 생성하며, draft가 얼마나 확장될 수 있는지 제한합니다. 한편, target verification은 첫 번째 거부를 넘어선 위치에서도 숨겨진 상태를 계산하지만, 이러한 상태는 일반적으로 T=0의 샘플링 온도보다 버려집니다. 우리는 ReAnchor를 제안합니다. ReAnchor는 Gemma 4-12B에서 가장 큰 평균 이익을 달성하여 19.4%의 수락 길이를 향상시키고 벤치마크를 통해 17.9%의 디코딩 속도를 향상시킵니다. 이 블록은 원래 블록 내의 예측을 향

Machine translation (Meta M2M100); the English original follows.

Abstract

Block diffusion drafters accelerate speculative decoding by generating multiple draft tokens in a single forward pass. However, each pass produces only a fixed-length block from a single target-generated anchor, limiting how far the draft can extend. Meanwhile, target verification computes hidden states even at positions beyond the first rejection, but these states are typically discarded rather than reused. We propose ReAnchor, a training-free method that reanchors a pretrained drafter at shifted positions using discarded verifier states to generate parallel draft blocks. These blocks improve predictions within the original block and extend coverage beyond its endpoint. At sampling temperature of T=0, ReAnchor achieves its largest average gains on Gemma 4-12B, improving acceptance length by 19.4% and decoding speed by 17.9% across eight benchmarks. Averaged equally over four target models and eight benchmarks, it improves acceptance length by 15.2% and decoding speed by 13.4% with pretrained DFlash drafters.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Decoding methods, Computer science, Block (permutation group theory), Reuse, Limiting, Algorithm

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange