Sanghyun Lee, Wonjin Nam, Seungho Jang, Jaeho Lee, Jiho Park, Jaewoon Lee, Seungryong Kim · 2026 · arXiv
DOI: 10.48550/arXiv.2610.26531 · open-access PDF
Drafter difusi blok mempercepat decoding spekulatif dengan menghasilkan token draft yang lebih banyak dalam satu forward pass. Namun, setiap pass hanya menghasilkan blok panjang tetap dari satu anchor yang dihasilkan oleh target tunggal, membatasi sejauh mana draft dapat diperluas. Sementara itu, verifikasi target menghitung keadaan tersembunyi bahkan di posisi di luar penolakan pertama, tetapi keadaan ini biasanya dibuang daripada digunakan kembali. Kami mengusulkan ReAnchor, metode bebas pelatihan yang merangkai kembali draft yang diprediksi pada posisi yang dipindahkan menggunakan keadaan verifier yang dibuang untuk menghasilkan blok draft paralel.
Machine translation (Meta M2M100); the English original follows.
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
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.26531, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange