{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:6SL6MBHYXCVJRN2R4TBM2TR7XY","short_pith_number":"pith:6SL6MBHY","schema_version":"1.0","canonical_sha256":"f497e604f8b8aa98b751e4c2cd4e3fbe14891120a631e29626cd696073ff50c1","source":{"kind":"arxiv","id":"2510.05688","version":2},"attestation_state":"computed","paper":{"title":"vAttention: Verified Sparse Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Aditya Desai, Alejandro Cuadron, Ion Stoica, Joseph E. Gonzalez, Kumar Krishna Agrawal, Luis Gaspar Schroeder, Matei Zaharia, Shuo Yang","submitted_at":"2025-10-07T08:46:08Z","abstract_excerpt":"State-of-the-art sparse attention methods for reducing decoding latency fall into two main categories: approximate top-$k$ (and its extension, top-$p$) and recently introduced sampling-based estimation. However, these approaches are fundamentally limited in their ability to approximate full attention: they fail to provide consistent approximations across heads and query vectors and, most critically, lack guarantees on approximation quality, limiting their practical deployment. We observe that top-$k$ and random sampling are complementary: top-$k$ performs well when attention scores are dominat"},"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":"2510.05688","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-10-07T08:46:08Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bb33cef3cd76d74abc9e9a3dc0ee425d070dffae611cf1fbfc3b7b98d2cea312","abstract_canon_sha256":"ef2cf657690db8a9283dcf50ef24be58859b682fd38b21b728567e96f90a9676"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:05:02.919661Z","signature_b64":"3epvlPuHqlQ7lQpvJt2tR/jjZFr1C9xSwZBiw88IUyho0Z2mlW6VX4CioTHknJhwXSdD4vyV0rm7kMj4sKbkDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f497e604f8b8aa98b751e4c2cd4e3fbe14891120a631e29626cd696073ff50c1","last_reissued_at":"2026-05-26T02:05:02.918665Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:05:02.918665Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"vAttention: Verified Sparse Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Aditya Desai, Alejandro Cuadron, Ion Stoica, Joseph E. Gonzalez, Kumar Krishna Agrawal, Luis Gaspar Schroeder, Matei Zaharia, Shuo Yang","submitted_at":"2025-10-07T08:46:08Z","abstract_excerpt":"State-of-the-art sparse attention methods for reducing decoding latency fall into two main categories: approximate top-$k$ (and its extension, top-$p$) and recently introduced sampling-based estimation. However, these approaches are fundamentally limited in their ability to approximate full attention: they fail to provide consistent approximations across heads and query vectors and, most critically, lack guarantees on approximation quality, limiting their practical deployment. We observe that top-$k$ and random sampling are complementary: top-$k$ performs well when attention scores are dominat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.05688","kind":"arxiv","version":2},"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/2510.05688/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":"2510.05688","created_at":"2026-05-26T02:05:02.918795+00:00"},{"alias_kind":"arxiv_version","alias_value":"2510.05688v2","created_at":"2026-05-26T02:05:02.918795+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.05688","created_at":"2026-05-26T02:05:02.918795+00:00"},{"alias_kind":"pith_short_12","alias_value":"6SL6MBHYXCVJ","created_at":"2026-05-26T02:05:02.918795+00:00"},{"alias_kind":"pith_short_16","alias_value":"6SL6MBHYXCVJRN2R","created_at":"2026-05-26T02:05:02.918795+00:00"},{"alias_kind":"pith_short_8","alias_value":"6SL6MBHY","created_at":"2026-05-26T02:05:02.918795+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2602.06283","citing_title":"SOCKET: SOft Collision Kernel EsTimator for Sparse Attention","ref_index":12,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6SL6MBHYXCVJRN2R4TBM2TR7XY","json":"https://pith.science/pith/6SL6MBHYXCVJRN2R4TBM2TR7XY.json","graph_json":"https://pith.science/api/pith-number/6SL6MBHYXCVJRN2R4TBM2TR7XY/graph.json","events_json":"https://pith.science/api/pith-number/6SL6MBHYXCVJRN2R4TBM2TR7XY/events.json","paper":"https://pith.science/paper/6SL6MBHY"},"agent_actions":{"view_html":"https://pith.science/pith/6SL6MBHYXCVJRN2R4TBM2TR7XY","download_json":"https://pith.science/pith/6SL6MBHYXCVJRN2R4TBM2TR7XY.json","view_paper":"https://pith.science/paper/6SL6MBHY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2510.05688&json=true","fetch_graph":"https://pith.science/api/pith-number/6SL6MBHYXCVJRN2R4TBM2TR7XY/graph.json","fetch_events":"https://pith.science/api/pith-number/6SL6MBHYXCVJRN2R4TBM2TR7XY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6SL6MBHYXCVJRN2R4TBM2TR7XY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6SL6MBHYXCVJRN2R4TBM2TR7XY/action/storage_attestation","attest_author":"https://pith.science/pith/6SL6MBHYXCVJRN2R4TBM2TR7XY/action/author_attestation","sign_citation":"https://pith.science/pith/6SL6MBHYXCVJRN2R4TBM2TR7XY/action/citation_signature","submit_replication":"https://pith.science/pith/6SL6MBHYXCVJRN2R4TBM2TR7XY/action/replication_record"}},"created_at":"2026-05-26T02:05:02.918795+00:00","updated_at":"2026-05-26T02:05:02.918795+00:00"}