{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UWCSRGJM4OSAZQZMCANZ3G2JMH","short_pith_number":"pith:UWCSRGJM","schema_version":"1.0","canonical_sha256":"a58528992ce3a40cc32c101b9d9b4961f7a78d58e483068fcc6600d0170f685d","source":{"kind":"arxiv","id":"2608.02901","version":1},"attestation_state":"computed","paper":{"title":"AnchorKV: Anchor-Residual KV Cache Compression","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Assaf Schuster, Malik Khalaf, Nitzan Hodos, Yara Shamshoum, Yuval Sieradzki","submitted_at":"2026-08-03T21:38:30Z","abstract_excerpt":"The key-value (KV) cache is the primary memory bottleneck in long-context LLM inference. Existing approaches attack it from opposite ends: eviction methods permanently discard tokens, degrading performance whenever a discarded token later proves essential, while quantization methods retain all tokens at low precision but offer limited compression. We propose AnchorKV, a compression scheme that shrinks the cache by $20\\times$ without discarding a single token. AnchorKV represents the cache using a small set of anchors stored exactly, expresses every other token through its most similar anchor, "},"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":"2608.02901","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-08-03T21:38:30Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"bdf355ad8041844b03295230db89e4dc7b268bfa95d43d4de2fc5c7765a349ef","abstract_canon_sha256":"535cfbc68464a1017cf3cc1c0377c7b4c7a22693116942ece6d44fa548fba007"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-05T00:43:34.177163Z","signature_b64":"avX7WDygvN2J8hsMIwhNdp4oqXE7NqSnjudlk3PDz61XsrZ14En5mW0jec95U11+VGiof4E/zgDfjdP05LsSDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a58528992ce3a40cc32c101b9d9b4961f7a78d58e483068fcc6600d0170f685d","last_reissued_at":"2026-08-05T00:43:34.174639Z","signature_status":"signed_v1","first_computed_at":"2026-08-05T00:43:34.174639Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AnchorKV: Anchor-Residual KV Cache Compression","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Assaf Schuster, Malik Khalaf, Nitzan Hodos, Yara Shamshoum, Yuval Sieradzki","submitted_at":"2026-08-03T21:38:30Z","abstract_excerpt":"The key-value (KV) cache is the primary memory bottleneck in long-context LLM inference. Existing approaches attack it from opposite ends: eviction methods permanently discard tokens, degrading performance whenever a discarded token later proves essential, while quantization methods retain all tokens at low precision but offer limited compression. We propose AnchorKV, a compression scheme that shrinks the cache by $20\\times$ without discarding a single token. AnchorKV represents the cache using a small set of anchors stored exactly, expresses every other token through its most similar anchor, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.02901","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/2608.02901/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":"2608.02901","created_at":"2026-08-05T00:43:34.175612+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.02901v1","created_at":"2026-08-05T00:43:34.175612+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.02901","created_at":"2026-08-05T00:43:34.175612+00:00"},{"alias_kind":"pith_short_12","alias_value":"UWCSRGJM4OSA","created_at":"2026-08-05T00:43:34.175612+00:00"},{"alias_kind":"pith_short_16","alias_value":"UWCSRGJM4OSAZQZM","created_at":"2026-08-05T00:43:34.175612+00:00"},{"alias_kind":"pith_short_8","alias_value":"UWCSRGJM","created_at":"2026-08-05T00:43:34.175612+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/UWCSRGJM4OSAZQZMCANZ3G2JMH","json":"https://pith.science/pith/UWCSRGJM4OSAZQZMCANZ3G2JMH.json","graph_json":"https://pith.science/api/pith-number/UWCSRGJM4OSAZQZMCANZ3G2JMH/graph.json","events_json":"https://pith.science/api/pith-number/UWCSRGJM4OSAZQZMCANZ3G2JMH/events.json","paper":"https://pith.science/paper/UWCSRGJM"},"agent_actions":{"view_html":"https://pith.science/pith/UWCSRGJM4OSAZQZMCANZ3G2JMH","download_json":"https://pith.science/pith/UWCSRGJM4OSAZQZMCANZ3G2JMH.json","view_paper":"https://pith.science/paper/UWCSRGJM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.02901&json=true","fetch_graph":"https://pith.science/api/pith-number/UWCSRGJM4OSAZQZMCANZ3G2JMH/graph.json","fetch_events":"https://pith.science/api/pith-number/UWCSRGJM4OSAZQZMCANZ3G2JMH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UWCSRGJM4OSAZQZMCANZ3G2JMH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UWCSRGJM4OSAZQZMCANZ3G2JMH/action/storage_attestation","attest_author":"https://pith.science/pith/UWCSRGJM4OSAZQZMCANZ3G2JMH/action/author_attestation","sign_citation":"https://pith.science/pith/UWCSRGJM4OSAZQZMCANZ3G2JMH/action/citation_signature","submit_replication":"https://pith.science/pith/UWCSRGJM4OSAZQZMCANZ3G2JMH/action/replication_record"}},"created_at":"2026-08-05T00:43:34.175612+00:00","updated_at":"2026-08-05T00:43:34.175612+00:00"}