{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:B2WWDVP4EFTOIHL6ZN37ROJQE2","short_pith_number":"pith:B2WWDVP4","schema_version":"1.0","canonical_sha256":"0ead61d5fc2166e41d7ecb77f8b93026aff9e48133b9436d2214f66a796db177","source":{"kind":"arxiv","id":"2601.20107","version":2},"attestation_state":"computed","paper":{"title":"Structural Anchor Pruning: Training-Free Multi-Vector Compression for Visual Document Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.IR"],"primary_cat":"cs.CV","authors_text":"Yao Zhang, Yu Xiao, Zhuchenyang Liu, Ziyu Hu","submitted_at":"2026-01-27T22:50:11Z","abstract_excerpt":"Recent Vision-Language Models (e.g., ColPali) enable fine-grained Visual Document Retrieval (VDR) but incur prohibitive multi-vector index storage overhead. Existing training-free pruning methods either rely on heuristic layer choices or degrade sharply under aggressive compression, leading prior work to argue that effective high-compression pruning requires query-dependent training. We challenge this view with Structural Anchor Pruning (SAP), a self-calibrating, training-free, and query-agnostic index-time pruning framework with three components: (i) Score Retention (SR), a white-box per-laye"},"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":"2601.20107","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-27T22:50:11Z","cross_cats_sorted":["cs.CL","cs.IR"],"title_canon_sha256":"41041bea6b1d6f3b7b8c8bb93f3bf8b238e1addba884856f64dd9fdc7f8e5486","abstract_canon_sha256":"13436cdb0cbef53aad774b4d678bebf353202ff58634dd451329529ee0cd9f1f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T01:03:54.716220Z","signature_b64":"gPWBXgnhXM1btYjmkpa4ebpY1dDEY1o3UmCDMz+D8Yp2Q4s7WOlbPh8ZaA0xqhglwgha0XzjuYxNOeDozgV9Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0ead61d5fc2166e41d7ecb77f8b93026aff9e48133b9436d2214f66a796db177","last_reissued_at":"2026-05-22T01:03:54.715292Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T01:03:54.715292Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Structural Anchor Pruning: Training-Free Multi-Vector Compression for Visual Document Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.IR"],"primary_cat":"cs.CV","authors_text":"Yao Zhang, Yu Xiao, Zhuchenyang Liu, Ziyu Hu","submitted_at":"2026-01-27T22:50:11Z","abstract_excerpt":"Recent Vision-Language Models (e.g., ColPali) enable fine-grained Visual Document Retrieval (VDR) but incur prohibitive multi-vector index storage overhead. Existing training-free pruning methods either rely on heuristic layer choices or degrade sharply under aggressive compression, leading prior work to argue that effective high-compression pruning requires query-dependent training. We challenge this view with Structural Anchor Pruning (SAP), a self-calibrating, training-free, and query-agnostic index-time pruning framework with three components: (i) Score Retention (SR), a white-box per-laye"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.20107","kind":"arxiv","version":2},"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/2601.20107/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":"2601.20107","created_at":"2026-05-22T01:03:54.715442+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.20107v2","created_at":"2026-05-22T01:03:54.715442+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.20107","created_at":"2026-05-22T01:03:54.715442+00:00"},{"alias_kind":"pith_short_12","alias_value":"B2WWDVP4EFTO","created_at":"2026-05-22T01:03:54.715442+00:00"},{"alias_kind":"pith_short_16","alias_value":"B2WWDVP4EFTOIHL6","created_at":"2026-05-22T01:03:54.715442+00:00"},{"alias_kind":"pith_short_8","alias_value":"B2WWDVP4","created_at":"2026-05-22T01:03:54.715442+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/B2WWDVP4EFTOIHL6ZN37ROJQE2","json":"https://pith.science/pith/B2WWDVP4EFTOIHL6ZN37ROJQE2.json","graph_json":"https://pith.science/api/pith-number/B2WWDVP4EFTOIHL6ZN37ROJQE2/graph.json","events_json":"https://pith.science/api/pith-number/B2WWDVP4EFTOIHL6ZN37ROJQE2/events.json","paper":"https://pith.science/paper/B2WWDVP4"},"agent_actions":{"view_html":"https://pith.science/pith/B2WWDVP4EFTOIHL6ZN37ROJQE2","download_json":"https://pith.science/pith/B2WWDVP4EFTOIHL6ZN37ROJQE2.json","view_paper":"https://pith.science/paper/B2WWDVP4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.20107&json=true","fetch_graph":"https://pith.science/api/pith-number/B2WWDVP4EFTOIHL6ZN37ROJQE2/graph.json","fetch_events":"https://pith.science/api/pith-number/B2WWDVP4EFTOIHL6ZN37ROJQE2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/B2WWDVP4EFTOIHL6ZN37ROJQE2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/B2WWDVP4EFTOIHL6ZN37ROJQE2/action/storage_attestation","attest_author":"https://pith.science/pith/B2WWDVP4EFTOIHL6ZN37ROJQE2/action/author_attestation","sign_citation":"https://pith.science/pith/B2WWDVP4EFTOIHL6ZN37ROJQE2/action/citation_signature","submit_replication":"https://pith.science/pith/B2WWDVP4EFTOIHL6ZN37ROJQE2/action/replication_record"}},"created_at":"2026-05-22T01:03:54.715442+00:00","updated_at":"2026-05-22T01:03:54.715442+00:00"}