{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:LDXXKYC4SALQ5OVRLBTKMCB2AX","short_pith_number":"pith:LDXXKYC4","schema_version":"1.0","canonical_sha256":"58ef75605c90170ebab15866a6083a05fed3282b0b826bf15f4641f652c82cbd","source":{"kind":"arxiv","id":"2604.12376","version":2},"attestation_state":"computed","paper":{"title":"Cooperative Memory Paging with Keyword Bookmarks for Long-Horizon LLM Conversations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Cooperative paging with keyword bookmarks lets LLMs recover evicted conversation turns on demand and achieve higher answer quality than full context or retrieval methods.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ziyang Liu","submitted_at":"2026-04-14T07:06:35Z","abstract_excerpt":"When LLM conversations grow beyond the context window, old content must be evicted -- but how does the model recover it when needed? We propose cooperative paging: evicted segments are replaced with minimal keyword bookmarks ([pN:keywords], ~8-24 tokens each), and the model is given a recall() tool to retrieve full content on demand. On the LoCoMo benchmark (10 real multi-session conversations, 300+ turns), cooperative paging achieves the highest answer quality among six methods -- outperforming truncation, BM25, word-overlap retrieval, a search-tool baseline, and full context -- on four model"},"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":"2604.12376","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-14T07:06:35Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1112b36fba36114aa4f805d0134c960d4070789b98d9f8cf6cb087abbb395df3","abstract_canon_sha256":"683db263687a16b74e6b0df378cbf99218c0ebca88c663919b5e74c13a8a4ec7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:03:30.017246Z","signature_b64":"e8H1Ex1uePBjBtjJYZeGpIVseKquwKvUAJQhzoEGWDYB5lcK+dS8VWqK4r7yWzS4xzvg4Gt12gUCFejRbZm+Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"58ef75605c90170ebab15866a6083a05fed3282b0b826bf15f4641f652c82cbd","last_reissued_at":"2026-05-26T01:03:30.016508Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:03:30.016508Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Cooperative Memory Paging with Keyword Bookmarks for Long-Horizon LLM Conversations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Cooperative paging with keyword bookmarks lets LLMs recover evicted conversation turns on demand and achieve higher answer quality than full context or retrieval methods.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ziyang Liu","submitted_at":"2026-04-14T07:06:35Z","abstract_excerpt":"When LLM conversations grow beyond the context window, old content must be evicted -- but how does the model recover it when needed? We propose cooperative paging: evicted segments are replaced with minimal keyword bookmarks ([pN:keywords], ~8-24 tokens each), and the model is given a recall() tool to retrieve full content on demand. On the LoCoMo benchmark (10 real multi-session conversations, 300+ turns), cooperative paging achieves the highest answer quality among six methods -- outperforming truncation, BM25, word-overlap retrieval, a search-tool baseline, and full context -- on four model"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"On the LoCoMo benchmark (10 real multi-session conversations, 300+ turns), cooperative paging achieves the highest answer quality among six methods -- outperforming truncation, BM25, word-overlap retrieval, a search-tool baseline, and full context -- on four models (GPT-4o-mini, DeepSeek-v3.2, Claude Haiku, GLM-5), confirmed by four independent LLM judges (p=0.017, paired bootstrap).","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That keyword bookmarks supply enough distinctive information for the model to correctly trigger and select the right page via the recall() tool, given the reported 57% correct selection rate when bookmarks lack specificity.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Cooperative paging replaces evicted LLM context with keyword bookmarks and adds a recall tool, outperforming six other methods on the LoCoMo benchmark across four models with statistical significance.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Cooperative paging with keyword bookmarks lets LLMs recover evicted conversation turns on demand and achieve higher answer quality than full context or retrieval methods.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"479463706bf92a7e932eec911d8ad1253953ea1607bfa88100b67ac76263762d"},"source":{"id":"2604.12376","kind":"arxiv","version":2},"verdict":{"id":"57840bdd-8026-4dee-9606-963b51535503","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T15:35:26.791767Z","strongest_claim":"On the LoCoMo benchmark (10 real multi-session conversations, 300+ turns), cooperative paging achieves the highest answer quality among six methods -- outperforming truncation, BM25, word-overlap retrieval, a search-tool baseline, and full context -- on four models (GPT-4o-mini, DeepSeek-v3.2, Claude Haiku, GLM-5), confirmed by four independent LLM judges (p=0.017, paired bootstrap).","one_line_summary":"Cooperative paging replaces evicted LLM context with keyword bookmarks and adds a recall tool, outperforming six other methods on the LoCoMo benchmark across four models with statistical significance.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That keyword bookmarks supply enough distinctive information for the model to correctly trigger and select the right page via the recall() tool, given the reported 57% correct selection rate when bookmarks lack specificity.","pith_extraction_headline":"Cooperative paging with keyword bookmarks lets LLMs recover evicted conversation turns on demand and achieve higher answer quality than full context or retrieval methods."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.12376/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":"2604.12376","created_at":"2026-05-26T01:03:30.016621+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.12376v2","created_at":"2026-05-26T01:03:30.016621+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.12376","created_at":"2026-05-26T01:03:30.016621+00:00"},{"alias_kind":"pith_short_12","alias_value":"LDXXKYC4SALQ","created_at":"2026-05-26T01:03:30.016621+00:00"},{"alias_kind":"pith_short_16","alias_value":"LDXXKYC4SALQ5OVR","created_at":"2026-05-26T01:03:30.016621+00:00"},{"alias_kind":"pith_short_8","alias_value":"LDXXKYC4","created_at":"2026-05-26T01:03:30.016621+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/LDXXKYC4SALQ5OVRLBTKMCB2AX","json":"https://pith.science/pith/LDXXKYC4SALQ5OVRLBTKMCB2AX.json","graph_json":"https://pith.science/api/pith-number/LDXXKYC4SALQ5OVRLBTKMCB2AX/graph.json","events_json":"https://pith.science/api/pith-number/LDXXKYC4SALQ5OVRLBTKMCB2AX/events.json","paper":"https://pith.science/paper/LDXXKYC4"},"agent_actions":{"view_html":"https://pith.science/pith/LDXXKYC4SALQ5OVRLBTKMCB2AX","download_json":"https://pith.science/pith/LDXXKYC4SALQ5OVRLBTKMCB2AX.json","view_paper":"https://pith.science/paper/LDXXKYC4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.12376&json=true","fetch_graph":"https://pith.science/api/pith-number/LDXXKYC4SALQ5OVRLBTKMCB2AX/graph.json","fetch_events":"https://pith.science/api/pith-number/LDXXKYC4SALQ5OVRLBTKMCB2AX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LDXXKYC4SALQ5OVRLBTKMCB2AX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LDXXKYC4SALQ5OVRLBTKMCB2AX/action/storage_attestation","attest_author":"https://pith.science/pith/LDXXKYC4SALQ5OVRLBTKMCB2AX/action/author_attestation","sign_citation":"https://pith.science/pith/LDXXKYC4SALQ5OVRLBTKMCB2AX/action/citation_signature","submit_replication":"https://pith.science/pith/LDXXKYC4SALQ5OVRLBTKMCB2AX/action/replication_record"}},"created_at":"2026-05-26T01:03:30.016621+00:00","updated_at":"2026-05-26T01:03:30.016621+00:00"}