跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.22191 · v1

RISA: Recoverable Sparse Attention for Block-Diffusion Language Models

首次提交 2026-11-17 12:27
版本更新 2026-11-17 12:27
雷达观测 2026-11-18 19:43
arXiv ↗PDF ↗

摘要

Block-wise diffusion language models (DLMs) refine multiple tokens in parallel, offering an alternative to token-by-token autoregressive decoding. Their long-context inference, however, remains bottlenecked by memory-bound attention: every denoising step repeatedly reads a growing historical key--value (KV) cache. Directly applying sparse attention does not fully solve this problem, because different queries can select different pages, inflating the block's collective read set, while repeated support construction adds overhead and permanent eviction can discard information needed later. We present RISA, a recoverable sparse-attention method with shared read plans for block-wise DLMs. RISA exploits the fact that historical KV remains fixed within a block while queries evolve across denoising steps. Newly generated history is organized into exact priority and recovery tiers; seed queries then build a protected, bounded read plan whose indices are reused across the block while attention is recomputed with fresh queries and current-block KV. This separates history preservation from repeated reading and amortizes sparse-support construction without changing the backbone or decoder. On DreamReasoner-8B, RISA retains comparable task performance. At 32K, the default configuration speeds up attention by 2.67×2.67\times and fixed-trajectory requests by 1.36×1.36\times over dense, and attention speedup reaches 5.53×5.53\times at 128K. Across eight settings on three models, RISA decodes faster than the tested MAGE, QUEST, and LoSA budgets at matched KL.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-17 12:27 · 最新观测版本

RISA: Recoverable Sparse Attention for Block-Diffusion Language Models

Block-wise diffusion language models (DLMs) refine multiple tokens in parallel, offering an alternative to token-by-token autoregressive decoding. Their long-context inference, however, remains bottlenecked by memory-bound attention: every denoising step repeatedly reads a growing historical key--value (KV) cache. Directly applying sparse attention does not fully solve this problem, because different queries can select different pages, inflating the block's collective read set, while repeated support construction adds overhead and permanent eviction can discard information needed later. We present RISA, a recoverable sparse-attention method with shared read plans for block-wise DLMs. RISA exploits the fact that historical KV remains fixed within a block while queries evolve across denoising steps. Newly generated history is organized into exact priority and recovery tiers; seed queries then build a protected, bounded read plan whose indices are reused across the block while attention is recomputed with fresh queries and current-block KV. This separates history preservation from repeated reading and amortizes sparse-support construction without changing the backbone or decoder. On DreamReasoner-8B, RISA retains comparable task performance. At 32K, the default configuration speeds up attention by 2.67×2.67\times and fixed-trajectory requests by 1.36×1.36\times over dense, and attention speedup reaches 5.53×5.53\times at 128K. Across eight settings on three models, RISA decodes faster than the tested MAGE, QUEST, and LoSA budgets at matched KL.

查看此版本解读 →查看此版本原文 ↗