{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RS72TYM2XIQVIOL57WQFKWG6CB","short_pith_number":"pith:RS72TYM2","schema_version":"1.0","canonical_sha256":"8cbfa9e19aba2154397dfda05558de105b2672a91e48c80dc64dae7a58a61396","source":{"kind":"arxiv","id":"2505.21919","version":1},"attestation_state":"computed","paper":{"title":"Towards Efficient Key-Value Cache Management for Prefix Prefilling in LLM Inference","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.DC"],"primary_cat":"cs.ET","authors_text":"Chen Wang, Eun Kyung Lee, Hao Yu, Yue Zhu, Zhuoran Liu","submitted_at":"2025-05-28T03:05:55Z","abstract_excerpt":"The increasing adoption of large language models (LLMs) with extended context windows necessitates efficient Key-Value Cache (KVC) management to optimize inference performance. Inference workloads like Retrieval-Augmented Generation (RAG) and agents exhibit high cache reusability, making efficient caching critical to reducing redundancy and improving speed. We analyze real-world KVC access patterns using publicly available traces and evaluate commercial key-value stores like Redis and state-of-the-art RDMA-based systems (CHIME [1] and Sherman [2]) for KVC metadata management. Our work demonstr"},"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":"2505.21919","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.ET","submitted_at":"2025-05-28T03:05:55Z","cross_cats_sorted":["cs.AI","cs.DC"],"title_canon_sha256":"2364870be6abd3471aeda8ca0a5528d94fe6658c7746bdf3dd1d46875666ec20","abstract_canon_sha256":"c65645de4fdb76691b1ba9435d83ff1f287a6052b34d3d73aa32cb2c823cae16"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:05.956291Z","signature_b64":"lWvEUALLAmktmgvFST9BtN+MWU9CMjtr9huHGZ0shNN+25l7fweVzMyiKSaL3A67Rex5uWRYaIWgHqry97WIDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8cbfa9e19aba2154397dfda05558de105b2672a91e48c80dc64dae7a58a61396","last_reissued_at":"2026-07-05T11:11:05.955768Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:05.955768Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Efficient Key-Value Cache Management for Prefix Prefilling in LLM Inference","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.DC"],"primary_cat":"cs.ET","authors_text":"Chen Wang, Eun Kyung Lee, Hao Yu, Yue Zhu, Zhuoran Liu","submitted_at":"2025-05-28T03:05:55Z","abstract_excerpt":"The increasing adoption of large language models (LLMs) with extended context windows necessitates efficient Key-Value Cache (KVC) management to optimize inference performance. Inference workloads like Retrieval-Augmented Generation (RAG) and agents exhibit high cache reusability, making efficient caching critical to reducing redundancy and improving speed. We analyze real-world KVC access patterns using publicly available traces and evaluate commercial key-value stores like Redis and state-of-the-art RDMA-based systems (CHIME [1] and Sherman [2]) for KVC metadata management. Our work demonstr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.21919","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/2505.21919/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":"2505.21919","created_at":"2026-07-05T11:11:05.955845+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.21919v1","created_at":"2026-07-05T11:11:05.955845+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.21919","created_at":"2026-07-05T11:11:05.955845+00:00"},{"alias_kind":"pith_short_12","alias_value":"RS72TYM2XIQV","created_at":"2026-07-05T11:11:05.955845+00:00"},{"alias_kind":"pith_short_16","alias_value":"RS72TYM2XIQVIOL5","created_at":"2026-07-05T11:11:05.955845+00:00"},{"alias_kind":"pith_short_8","alias_value":"RS72TYM2","created_at":"2026-07-05T11:11:05.955845+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/RS72TYM2XIQVIOL57WQFKWG6CB","json":"https://pith.science/pith/RS72TYM2XIQVIOL57WQFKWG6CB.json","graph_json":"https://pith.science/api/pith-number/RS72TYM2XIQVIOL57WQFKWG6CB/graph.json","events_json":"https://pith.science/api/pith-number/RS72TYM2XIQVIOL57WQFKWG6CB/events.json","paper":"https://pith.science/paper/RS72TYM2"},"agent_actions":{"view_html":"https://pith.science/pith/RS72TYM2XIQVIOL57WQFKWG6CB","download_json":"https://pith.science/pith/RS72TYM2XIQVIOL57WQFKWG6CB.json","view_paper":"https://pith.science/paper/RS72TYM2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.21919&json=true","fetch_graph":"https://pith.science/api/pith-number/RS72TYM2XIQVIOL57WQFKWG6CB/graph.json","fetch_events":"https://pith.science/api/pith-number/RS72TYM2XIQVIOL57WQFKWG6CB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RS72TYM2XIQVIOL57WQFKWG6CB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RS72TYM2XIQVIOL57WQFKWG6CB/action/storage_attestation","attest_author":"https://pith.science/pith/RS72TYM2XIQVIOL57WQFKWG6CB/action/author_attestation","sign_citation":"https://pith.science/pith/RS72TYM2XIQVIOL57WQFKWG6CB/action/citation_signature","submit_replication":"https://pith.science/pith/RS72TYM2XIQVIOL57WQFKWG6CB/action/replication_record"}},"created_at":"2026-07-05T11:11:05.955845+00:00","updated_at":"2026-07-05T11:11:05.955845+00:00"}