The paper introduces Pre-hoc Sparsity (PrHS), a method that selects key-value (KV) cache entries before attention scoring to avoid posterior bias in large language model inference. By bounding mutual‑information loss through the dropped attention mass, PrHS offers explicit accuracy control and implements three orthogonal selectors across time, depth, and layer. Experiments on LLaMA and Mistral models show that PrHS cuts retrieval overhead by over 90%, achieves higher sparsity than HShare, and delivers significant speedups and reduced FLOPs on NVIDIA A100 GPUs while maintaining near‑dense accuracy.
By Yifei Gao, Lei Wang, Rong-Cheng Tu, Qixin Zhang, Jun Cheng, Dacheng Tao
arXiv:2609.36722v1 Announce Type: new
Abstract: Large language model (LLM) agents repeatedly load reusable content, such as skills, documents, and memory entries, into the current context. Re-encodin...
By Xinghao Chen, Junnan Dong, Cai Ke, Chak Tou Leong, Haocheng Sun, Keyu Chen, Siyu An, Ruizhi Qiao, Xing Sun, Wenjie Li, Xiaoyu Shen
arXiv:2606.02737v2 Announce Type: replace-cross
Abstract: Dense retrieval compresses a passage into a single vector, but this compression is positionally skewed: early content dominates the embedding...
By Andrianos Michail, Elias Schuhmacher, Juri Opitz, Simon Clematide, Rico Sennrich
arXiv:2606. 04302v1 Announce Type: cross Abstract: Key-value (KV) caching accelerates inference of large language models (LLMs) by reusing past computations for generated tokens.
By Haocheng Xia, Mihir Pamnani, Hanxi Fang, Supawit Chockchowwat, Yongjoo Park
Public institutions hold large volumes of sensitive documents and support tickets that cannot leave the premises, ruling out cloud-hosted language models entirely. We report on RAGAL, a retrieval-augmented assistant for the technical-support team of AFIR, the Romanian Agency for Financing Rural Investments, built and operated under three hard constraints: zero data egress (no external API calls, even for synthetic data), a read-only mandate (the assistant drafts, humans execute), and a single 8 GB consumer laptop as the only development and training machine.
arXiv:2607. 27692v1 Announce Type: cross Abstract: Top-$K$ sparse attention reduces the cost of Softmax and value aggregation by attending to only a small subset of key--value (KV) entries.
By Wenshuai Yao, Wenyong Zhou, Hanyong Shao, Yizhe Chen, Zhiyuan Ning, Yuannuo Feng, Ru Huang, Kechao Tang
arXiv:2606. 05875v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) improves large language model (LLM) answer quality by grounding generation in external evidence, but processing retrieved contexts makes the prefill stage a dominant serving cost.
By Jianxin Yan, Wangze Ni, Zhenxin Li, Jiabao Jin, Zhitao Shen, Haoyang Li, Jia Zhu, Peng Cheng, Xuemin Lin, Lei Chen, Kui Ren
arXiv:2609.05760v1 Announce Type: cross
Abstract: We present RAGMark, a modular benchmarking framework for advanced Retrieval-Augmented Generation (RAG) systems targeting small-scale multi-GPU enviro...
By Zlatan Feric, Amir Taherin, Bin Ren, Yanzhi Wang, Jennifer Dy, David Kaeli
arXiv:2604. 20920v2 Announce Type: replace Abstract: Sparse attention can reduce the cost of long-context inference, but most variants introduce new architectural components.
By Yuzhen Mao, Michael Y. Li, Emily B. Fox
Block Sparse Flash Attention (BSFA) is a drop‑in replacement for FlashAttention that speeds up long‑context inference by pruning about 50% of computation and memory transfers. It selects the top‑k most important value blocks for each query using exact query‑key similarities and calibrated per‑layer, per‑head thresholds, requiring only a one‑time training‑free calibration. On Llama‑3.1‑8B, BSFA delivers up to 1.13× speedup on LongBench with a 1.1% accuracy drop and up to 1.24× on Needle‑in‑a‑Haystack retrieval with a 1% drop, while the attention kernel itself accelerates by up to 1.38×.
By Daniel Ohayon, Itay Lamprecht, Itay Hubara, Israel Cohen, Daniel Soudry, Noam Elata
The paper introduces REVA, a method for compressing retrieval-augmented generation (RAG) prompts by aggregating historical query–document–model interactions into reusable evidence views. REVA mines attention traces from the target generator, maps token-level attention to readable words, aggregates importance across repeated document accesses, and produces budget‑specific plain‑text views that maintain document order and the standard RAG interface. Experiments on four benchmarks with modern LLMs show that REVA improves generation quality by 1.0–5.8 points over existing compressors while reducing compression overhead by 5.3 to 15.6 times and adding less than 40 ms of latency.
By Tuan Nguyen, Qiran Hu, Banruo Liu, Khoa D. Doan, Kok-Seng Wong, Fan Lai
arXiv:2606. 07703v1 Announce Type: cross Abstract: Long-context prefill remains expensive because full/GQA layers still score the historical sequence, even in hybrid models with local, sparse, linear, or recurrent components.
By Hongxing Wang, Harenome Razanajato, Zhen Zhang, Yujie Yuan, Hongsheng Liu