{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SHK7BCSQZMGLQIE53BMMXFKAFO","short_pith_number":"pith:SHK7BCSQ","schema_version":"1.0","canonical_sha256":"91d5f08a50cb0cb8209dd858cb95402b9e05cba81e8bc34c1223e1ae3a5a1699","source":{"kind":"arxiv","id":"2608.01662","version":1},"attestation_state":"computed","paper":{"title":"LongCat Sparse Attention: Taming the Lightning via Streaming-aware Hierarchical Cross-Layer Indexing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.DC","cs.LG"],"primary_cat":"cs.AI","authors_text":"Cunguang Wang, Duyue Ma, Fengcun Li, Guanyu Wu, Hong Liu, Jianchao Tan, Jiaqi Zhang, Peng Pei, Wen Zan, Xiang Li, Xunliang Cai, Yerui Sun, Yifan Lu, Yuchen Xie","submitted_at":"2026-08-03T03:51:21Z","abstract_excerpt":"DeepSeek Sparse Attention (DSA) enables efficient long-context modeling through its Lightning Indexer. However, practical deployment remains constrained by the indexer's expensive $O(L^2)$ scoring overhead and the hardware-inefficient, discontinuous memory-access patterns induced by its outputs. To address these system-level bottlenecks, we introduce LongCat Sparse Attention (LSA), a hardware-algorithm co-designed framework comprising three complementary and orthogonal strategies: (1) Streaming-Aware Indexing, which selectively converts scattered KV entries into hardware-aligned contiguous lay"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2608.01662","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-03T03:51:21Z","cross_cats_sorted":["cs.CL","cs.DC","cs.LG"],"title_canon_sha256":"ff2d4bbf46d5fe6fe71a1bbf6f0a78488352db75e589738a94496283e20ca064","abstract_canon_sha256":"340916f898cc1c09dc27cc1d83cd333adca798098871f05ea01b9561a5db5cde"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T02:06:21.979024Z","signature_b64":"h6k3BnnucmfN1T+DmlUrMypCZktBqLCm/C3Uwgk08DfSR40XrQiZW6nAwzNFQ7mc6u+5VhEmk0v0hhCQzAoQAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"91d5f08a50cb0cb8209dd858cb95402b9e05cba81e8bc34c1223e1ae3a5a1699","last_reissued_at":"2026-08-04T02:06:21.977479Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T02:06:21.977479Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LongCat Sparse Attention: Taming the Lightning via Streaming-aware Hierarchical Cross-Layer Indexing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.DC","cs.LG"],"primary_cat":"cs.AI","authors_text":"Cunguang Wang, Duyue Ma, Fengcun Li, Guanyu Wu, Hong Liu, Jianchao Tan, Jiaqi Zhang, Peng Pei, Wen Zan, Xiang Li, Xunliang Cai, Yerui Sun, Yifan Lu, Yuchen Xie","submitted_at":"2026-08-03T03:51:21Z","abstract_excerpt":"DeepSeek Sparse Attention (DSA) enables efficient long-context modeling through its Lightning Indexer. However, practical deployment remains constrained by the indexer's expensive $O(L^2)$ scoring overhead and the hardware-inefficient, discontinuous memory-access patterns induced by its outputs. To address these system-level bottlenecks, we introduce LongCat Sparse Attention (LSA), a hardware-algorithm co-designed framework comprising three complementary and orthogonal strategies: (1) Streaming-Aware Indexing, which selectively converts scattered KV entries into hardware-aligned contiguous lay"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.01662","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2608.01662/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2608.01662","created_at":"2026-08-04T02:06:21.978957+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.01662v1","created_at":"2026-08-04T02:06:21.978957+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.01662","created_at":"2026-08-04T02:06:21.978957+00:00"},{"alias_kind":"pith_short_12","alias_value":"SHK7BCSQZMGL","created_at":"2026-08-04T02:06:21.978957+00:00"},{"alias_kind":"pith_short_16","alias_value":"SHK7BCSQZMGLQIE5","created_at":"2026-08-04T02:06:21.978957+00:00"},{"alias_kind":"pith_short_8","alias_value":"SHK7BCSQ","created_at":"2026-08-04T02:06:21.978957+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SHK7BCSQZMGLQIE53BMMXFKAFO","json":"https://pith.science/pith/SHK7BCSQZMGLQIE53BMMXFKAFO.json","graph_json":"https://pith.science/api/pith-number/SHK7BCSQZMGLQIE53BMMXFKAFO/graph.json","events_json":"https://pith.science/api/pith-number/SHK7BCSQZMGLQIE53BMMXFKAFO/events.json","paper":"https://pith.science/paper/SHK7BCSQ"},"agent_actions":{"view_html":"https://pith.science/pith/SHK7BCSQZMGLQIE53BMMXFKAFO","download_json":"https://pith.science/pith/SHK7BCSQZMGLQIE53BMMXFKAFO.json","view_paper":"https://pith.science/paper/SHK7BCSQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.01662&json=true","fetch_graph":"https://pith.science/api/pith-number/SHK7BCSQZMGLQIE53BMMXFKAFO/graph.json","fetch_events":"https://pith.science/api/pith-number/SHK7BCSQZMGLQIE53BMMXFKAFO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SHK7BCSQZMGLQIE53BMMXFKAFO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SHK7BCSQZMGLQIE53BMMXFKAFO/action/storage_attestation","attest_author":"https://pith.science/pith/SHK7BCSQZMGLQIE53BMMXFKAFO/action/author_attestation","sign_citation":"https://pith.science/pith/SHK7BCSQZMGLQIE53BMMXFKAFO/action/citation_signature","submit_replication":"https://pith.science/pith/SHK7BCSQZMGLQIE53BMMXFKAFO/action/replication_record"}},"created_at":"2026-08-04T02:06:21.978957+00:00","updated_at":"2026-08-04T02:06:21.978957+00:00"}