The paper introduces Stratified Sparse Attention (SSA), a framework that blends deterministic top‑k retrieval with score‑guided sampling to reduce the gap between sparse and dense attention in large language models. SSA attends to high‑scoring tokens deterministically while sampling the remaining attention mass using proxy scores, thereby recovering more of the attention distribution. Experiments on benchmarks such as RULER, LongBench, LOFT, and AIME show that SSA outperforms existing sparse methods across models ranging from 3B to 27B parameters, achieving significant accuracy gains and up to 2.96× decode speedup with optimized CUDA kernels.
By Sahil Joshi, Prithvi Dixit, Aryan Katiyar, Agniva Chowdhury, Joseph E. Gonzalez, Anshumali Shrivastava, Aditya Desai, Kumar Krishna Agrawal
The paper investigates how attention dynamics evolve across recurrent depth in language models, finding that attention support stabilizes early while hidden states and outputs take longer. It proposes WISE, a training‑free method that uses full attention in early steps and then reuses the discovered sparse working set for later steps, preserving performance on multi‑hop QA tasks. Experiments show that WISE maintains quality up to 2K context, offers measurable speedups, and highlights the importance of recurrent discovery of attention support.
By Ke Wan, Chen Chen
arXiv:2605. 16928v2 Announce Type: replace-cross Abstract: Long-context inference in large language models is bottlenecked by the quadratic cost of full attention.
By Yanke Zhou, Yiduo Li, Hanlin Tang, Maohua Li, Kan Liu, Tao Lan, Lin Qu, Yuan Yao, Xiaoxing Ma
Prism Attention introduces a proposal‑refined sparse attention framework that replaces static cross‑layer index reuse with a coarse‑to‑fine strategy, allowing layer‑wise adaptability while maintaining efficiency. The method leverages the observation that high‑attention tokens remain largely stable across layers, with changes concentrated near the boundary of previously selected regions. Experiments demonstrate state‑of‑the‑art performance on reasoning and long‑context benchmarks, achieving a 3.4× decoding speedup over FlashAttention at 128K context length.
By Zhenxu Tian, Kebin Liu, Zhengwu Yang, Yi Su, Qingqing Dang, Yanlin Sha, Kaipeng Deng, Dianhai Yu, Yanjun Ma, Juntao Li, Min Zhang
Elastic Threshold Attention (ETA) is a trainable attention mechanism that dynamically predicts contextual thresholds from query representations, enabling selective pruning of KV cache tokens during long‑context decoding. By multiplicatively suppressing sub‑threshold logits during training, ETA avoids representation collapse and eliminates localized attention sinks, allowing a 1.45B model to match dense attention performance at roughly 85% training sparsity and 38% active decode density. At inference, a custom Triton kernel achieves up to 2.5× faster decoding on sequences up to 512K tokens, and an offline calibration step can further reduce compute by 27% by freezing per‑head thresholds.
By Themistoklis Haris, Henry Li, Maryam Karimzadehgan
arXiv:2608.29058v1 Announce Type: new
Abstract: Dynamic sparse attention can reduce the quadratic cost of long-context prefilling without changing model weights. MInference assigns each attention hea...
By Chao Zhang, Yifan Ji, Ziyan Zhang, Kai Song, Fei Lin