{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:CNPMIFD42NONUMCBYNP4XLPY53","short_pith_number":"pith:CNPMIFD4","schema_version":"1.0","canonical_sha256":"135ec4147cd35cda3041c35fcbadf8eec5d308d34d835d0bcc50c7d37f50b12c","source":{"kind":"arxiv","id":"2606.01751","version":1},"attestation_state":"computed","paper":{"title":"SparseX: Efficient Segment-Level KV Cache Sharing for Interleaved LLM Serving","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.PF","authors_text":"Bo Tang, Feiyu Xiong, Kai Chen, Ning Liao, Quqing Zhang, Xiaoxing Wang, Zehao Lin","submitted_at":"2026-06-01T06:12:55Z","abstract_excerpt":"In long-context LLM serving, the prefill stage often dominates time-to-first-token and computational cost. Although Prefix Cache in vLLM/PagedAttention has been widely used to reuse identical prompt prefixes, repeated content in practical applications frequently appears as non-prefix, cross-request, cross-turn, and cross-agent segments, which makes conventional cache mechanisms insufficient. This paper presents SparseX, a segment-level KV Cache sharing method for common serving scenarios. SparseX uses contiguous token segments as reuse units and exploits Sparse-Q indices that naturally arise i"},"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":"2606.01751","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.PF","submitted_at":"2026-06-01T06:12:55Z","cross_cats_sorted":[],"title_canon_sha256":"a8c4d602e0b285fd37e8d7aaef26d0a20e7140a4f93478918d9327aca634b1fa","abstract_canon_sha256":"16c4b88d2f47dc0770eca7fb5268d85bfc2073720d5a079d4b09b353a5923c8a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:55.726505Z","signature_b64":"G7oOpf0AXgcO6VSv0d5urb1L11d3CU7JhrAZMkt8q2gwGnULxr+GtV6cmk0veAfd+TXSHPPR3sjHNSVvGx4rAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"135ec4147cd35cda3041c35fcbadf8eec5d308d34d835d0bcc50c7d37f50b12c","last_reissued_at":"2026-06-02T02:04:55.726084Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:55.726084Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SparseX: Efficient Segment-Level KV Cache Sharing for Interleaved LLM Serving","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.PF","authors_text":"Bo Tang, Feiyu Xiong, Kai Chen, Ning Liao, Quqing Zhang, Xiaoxing Wang, Zehao Lin","submitted_at":"2026-06-01T06:12:55Z","abstract_excerpt":"In long-context LLM serving, the prefill stage often dominates time-to-first-token and computational cost. Although Prefix Cache in vLLM/PagedAttention has been widely used to reuse identical prompt prefixes, repeated content in practical applications frequently appears as non-prefix, cross-request, cross-turn, and cross-agent segments, which makes conventional cache mechanisms insufficient. This paper presents SparseX, a segment-level KV Cache sharing method for common serving scenarios. SparseX uses contiguous token segments as reuse units and exploits Sparse-Q indices that naturally arise i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.01751","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/2606.01751/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":"2606.01751","created_at":"2026-06-02T02:04:55.726149+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.01751v1","created_at":"2026-06-02T02:04:55.726149+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.01751","created_at":"2026-06-02T02:04:55.726149+00:00"},{"alias_kind":"pith_short_12","alias_value":"CNPMIFD42NON","created_at":"2026-06-02T02:04:55.726149+00:00"},{"alias_kind":"pith_short_16","alias_value":"CNPMIFD42NONUMCB","created_at":"2026-06-02T02:04:55.726149+00:00"},{"alias_kind":"pith_short_8","alias_value":"CNPMIFD4","created_at":"2026-06-02T02:04:55.726149+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/CNPMIFD42NONUMCBYNP4XLPY53","json":"https://pith.science/pith/CNPMIFD42NONUMCBYNP4XLPY53.json","graph_json":"https://pith.science/api/pith-number/CNPMIFD42NONUMCBYNP4XLPY53/graph.json","events_json":"https://pith.science/api/pith-number/CNPMIFD42NONUMCBYNP4XLPY53/events.json","paper":"https://pith.science/paper/CNPMIFD4"},"agent_actions":{"view_html":"https://pith.science/pith/CNPMIFD42NONUMCBYNP4XLPY53","download_json":"https://pith.science/pith/CNPMIFD42NONUMCBYNP4XLPY53.json","view_paper":"https://pith.science/paper/CNPMIFD4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.01751&json=true","fetch_graph":"https://pith.science/api/pith-number/CNPMIFD42NONUMCBYNP4XLPY53/graph.json","fetch_events":"https://pith.science/api/pith-number/CNPMIFD42NONUMCBYNP4XLPY53/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CNPMIFD42NONUMCBYNP4XLPY53/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CNPMIFD42NONUMCBYNP4XLPY53/action/storage_attestation","attest_author":"https://pith.science/pith/CNPMIFD42NONUMCBYNP4XLPY53/action/author_attestation","sign_citation":"https://pith.science/pith/CNPMIFD42NONUMCBYNP4XLPY53/action/citation_signature","submit_replication":"https://pith.science/pith/CNPMIFD42NONUMCBYNP4XLPY53/action/replication_record"}},"created_at":"2026-06-02T02:04:55.726149+00:00","updated_at":"2026-06-02T02:04:55.726149+00:00"}