{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:V3ELPYBATONKYMFIOZKIYRXNJH","short_pith_number":"pith:V3ELPYBA","schema_version":"1.0","canonical_sha256":"aec8b7e0209b9aac30a876548c46ed49dd4d0108699d06d21d0b24b822dc4fef","source":{"kind":"arxiv","id":"2608.01657","version":1},"attestation_state":"computed","paper":{"title":"Preserving Admission Responsibility in Multi-Tenant Large Language Model Prefix Caches","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Rajkumar Buyya, Zhiyu Wang","submitted_at":"2026-08-03T03:47:14Z","abstract_excerpt":"Shared prefix caching turns Graphics Processing Unit (GPU) memory into persistent state shared across Large Language Model (LLM) tenants. A group that materializes new Key-Value (KV) blocks can force another to lose reusable state, yet request-time schedulers account for transient service, replacement policies primarily rank object value, and static partitioning strands idle capacity. We call this mismatch the admission-responsibility gap. To close it, we propose PrefixShield, which meters newly materialized full KV blocks, carries responsibility across requests, gates reuse promotion while de"},"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":"2608.01657","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2026-08-03T03:47:14Z","cross_cats_sorted":[],"title_canon_sha256":"98c71147c6925be205fc5eeb8e8a88070dd4e72d6fc3169b3e11bdb0bf4278ff","abstract_canon_sha256":"1c52983cb318f839699884ca9304d2790fa56a612f9d1f5a2dc119958a7d590e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T02:06:21.260399Z","signature_b64":"ZOd86dO8cDBN7bGjEHa5VFY9A/lM6be/sBEGjc8AT4vFkJqciHbC1kroNZxYjYsWZTSQaiTii6vQRL1sttRAAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aec8b7e0209b9aac30a876548c46ed49dd4d0108699d06d21d0b24b822dc4fef","last_reissued_at":"2026-08-04T02:06:21.258793Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T02:06:21.258793Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Preserving Admission Responsibility in Multi-Tenant Large Language Model Prefix Caches","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Rajkumar Buyya, Zhiyu Wang","submitted_at":"2026-08-03T03:47:14Z","abstract_excerpt":"Shared prefix caching turns Graphics Processing Unit (GPU) memory into persistent state shared across Large Language Model (LLM) tenants. A group that materializes new Key-Value (KV) blocks can force another to lose reusable state, yet request-time schedulers account for transient service, replacement policies primarily rank object value, and static partitioning strands idle capacity. We call this mismatch the admission-responsibility gap. To close it, we propose PrefixShield, which meters newly materialized full KV blocks, carries responsibility across requests, gates reuse promotion while de"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.01657","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/2608.01657/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":"2608.01657","created_at":"2026-08-04T02:06:21.260329+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.01657v1","created_at":"2026-08-04T02:06:21.260329+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.01657","created_at":"2026-08-04T02:06:21.260329+00:00"},{"alias_kind":"pith_short_12","alias_value":"V3ELPYBATONK","created_at":"2026-08-04T02:06:21.260329+00:00"},{"alias_kind":"pith_short_16","alias_value":"V3ELPYBATONKYMFI","created_at":"2026-08-04T02:06:21.260329+00:00"},{"alias_kind":"pith_short_8","alias_value":"V3ELPYBA","created_at":"2026-08-04T02:06:21.260329+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/V3ELPYBATONKYMFIOZKIYRXNJH","json":"https://pith.science/pith/V3ELPYBATONKYMFIOZKIYRXNJH.json","graph_json":"https://pith.science/api/pith-number/V3ELPYBATONKYMFIOZKIYRXNJH/graph.json","events_json":"https://pith.science/api/pith-number/V3ELPYBATONKYMFIOZKIYRXNJH/events.json","paper":"https://pith.science/paper/V3ELPYBA"},"agent_actions":{"view_html":"https://pith.science/pith/V3ELPYBATONKYMFIOZKIYRXNJH","download_json":"https://pith.science/pith/V3ELPYBATONKYMFIOZKIYRXNJH.json","view_paper":"https://pith.science/paper/V3ELPYBA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.01657&json=true","fetch_graph":"https://pith.science/api/pith-number/V3ELPYBATONKYMFIOZKIYRXNJH/graph.json","fetch_events":"https://pith.science/api/pith-number/V3ELPYBATONKYMFIOZKIYRXNJH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V3ELPYBATONKYMFIOZKIYRXNJH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V3ELPYBATONKYMFIOZKIYRXNJH/action/storage_attestation","attest_author":"https://pith.science/pith/V3ELPYBATONKYMFIOZKIYRXNJH/action/author_attestation","sign_citation":"https://pith.science/pith/V3ELPYBATONKYMFIOZKIYRXNJH/action/citation_signature","submit_replication":"https://pith.science/pith/V3ELPYBATONKYMFIOZKIYRXNJH/action/replication_record"}},"created_at":"2026-08-04T02:06:21.260329+00:00","updated_at":"2026-08-04T02:06:21.260329+00:00"}