{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:GMAKPLEXG2XAJ6HQKA6DEC3IM4","short_pith_number":"pith:GMAKPLEX","schema_version":"1.0","canonical_sha256":"3300a7ac9736ae04f8f0503c320b6867384499cf6109a5c46557f0b2ee427a3f","source":{"kind":"arxiv","id":"2501.17788","version":3},"attestation_state":"computed","paper":{"title":"WARP: An Efficient Engine for Multi-Vector Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Christopher Potts, Gustavo Alonso, Jan Luca Scheerer, Matei Zaharia, Omar Khattab","submitted_at":"2025-01-29T17:26:47Z","abstract_excerpt":"Multi-vector retrieval methods such as ColBERT and its recent variant, the ConteXtualized Token Retriever (XTR), offer high accuracy but face efficiency challenges at scale. To address this, we present WARP, a retrieval engine that substantially improves the efficiency of retrievers trained with the XTR objective through three key innovations: (1) WARP$_\\text{SELECT}$ for dynamic similarity imputation; (2) implicit decompression, avoiding costly vector reconstruction during retrieval; and (3) a two-stage reduction process for efficient score aggregation. Combined with highly-optimized C++ kern"},"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":"2501.17788","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2025-01-29T17:26:47Z","cross_cats_sorted":[],"title_canon_sha256":"ca2a8697826343203b7864a3b56be3df31438838201ba2c3bc408ca9f16d6332","abstract_canon_sha256":"17df20974e4317b4a7532422ef2785e1a07e2196a823492ac1d7644ac9c67721"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:32:23.906604Z","signature_b64":"OCg2wZbOMpNNn9WkuPUfn0PwRUhb2/CsIp4Pe0AqSCUL0v/2nnv5R6VsCDT9AM65sc44zNpFTrDgNpYtgt2MBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3300a7ac9736ae04f8f0503c320b6867384499cf6109a5c46557f0b2ee427a3f","last_reissued_at":"2026-07-05T11:32:23.906092Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:32:23.906092Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"WARP: An Efficient Engine for Multi-Vector Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Christopher Potts, Gustavo Alonso, Jan Luca Scheerer, Matei Zaharia, Omar Khattab","submitted_at":"2025-01-29T17:26:47Z","abstract_excerpt":"Multi-vector retrieval methods such as ColBERT and its recent variant, the ConteXtualized Token Retriever (XTR), offer high accuracy but face efficiency challenges at scale. To address this, we present WARP, a retrieval engine that substantially improves the efficiency of retrievers trained with the XTR objective through three key innovations: (1) WARP$_\\text{SELECT}$ for dynamic similarity imputation; (2) implicit decompression, avoiding costly vector reconstruction during retrieval; and (3) a two-stage reduction process for efficient score aggregation. Combined with highly-optimized C++ kern"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.17788","kind":"arxiv","version":3},"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/2501.17788/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":"2501.17788","created_at":"2026-07-05T11:32:23.906152+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.17788v3","created_at":"2026-07-05T11:32:23.906152+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.17788","created_at":"2026-07-05T11:32:23.906152+00:00"},{"alias_kind":"pith_short_12","alias_value":"GMAKPLEXG2XA","created_at":"2026-07-05T11:32:23.906152+00:00"},{"alias_kind":"pith_short_16","alias_value":"GMAKPLEXG2XAJ6HQ","created_at":"2026-07-05T11:32:23.906152+00:00"},{"alias_kind":"pith_short_8","alias_value":"GMAKPLEX","created_at":"2026-07-05T11:32:23.906152+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/GMAKPLEXG2XAJ6HQKA6DEC3IM4","json":"https://pith.science/pith/GMAKPLEXG2XAJ6HQKA6DEC3IM4.json","graph_json":"https://pith.science/api/pith-number/GMAKPLEXG2XAJ6HQKA6DEC3IM4/graph.json","events_json":"https://pith.science/api/pith-number/GMAKPLEXG2XAJ6HQKA6DEC3IM4/events.json","paper":"https://pith.science/paper/GMAKPLEX"},"agent_actions":{"view_html":"https://pith.science/pith/GMAKPLEXG2XAJ6HQKA6DEC3IM4","download_json":"https://pith.science/pith/GMAKPLEXG2XAJ6HQKA6DEC3IM4.json","view_paper":"https://pith.science/paper/GMAKPLEX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.17788&json=true","fetch_graph":"https://pith.science/api/pith-number/GMAKPLEXG2XAJ6HQKA6DEC3IM4/graph.json","fetch_events":"https://pith.science/api/pith-number/GMAKPLEXG2XAJ6HQKA6DEC3IM4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GMAKPLEXG2XAJ6HQKA6DEC3IM4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GMAKPLEXG2XAJ6HQKA6DEC3IM4/action/storage_attestation","attest_author":"https://pith.science/pith/GMAKPLEXG2XAJ6HQKA6DEC3IM4/action/author_attestation","sign_citation":"https://pith.science/pith/GMAKPLEXG2XAJ6HQKA6DEC3IM4/action/citation_signature","submit_replication":"https://pith.science/pith/GMAKPLEXG2XAJ6HQKA6DEC3IM4/action/replication_record"}},"created_at":"2026-07-05T11:32:23.906152+00:00","updated_at":"2026-07-05T11:32:23.906152+00:00"}