IndexCache: Accelerating Sparse Attention via Cross-Layer Index Reuse · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2603.12201/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTIndexCache: Accelerating Sparse Attention via Cross-Layer Index Reuse — mentioned by teleforcehttps://news.ycombinator.com/item?id=48569213urn:hackerpapers:mention:2603.12201:48569213Wed, 17 Jun 2026 12:05:15 GMT<p>&gt;This repository provides a patch for SGLang and vLLM that enables IndexCache inference acceleration for models using DeepSeek Sparse Attention (DSA), including DeepSeek-V3.2 and GLM-5. Paper here [1]. [1] IndexCache: Accelerating Sparse Attention via Cross-Layer Index Reuse: https://arxiv.org/abs/2603.12201</p><p><a href="https://hackerpaper.kaipfister.de/p/2603.12201/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.CL