{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:46OIZDUYOU4MBUQRLOGH6UNIMB","short_pith_number":"pith:46OIZDUY","schema_version":"1.0","canonical_sha256":"e79c8c8e987538c0d2115b8c7f51a8605efea5c415f23b924d6bce11663b5675","source":{"kind":"arxiv","id":"2608.10636","version":1},"attestation_state":"computed","paper":{"title":"DistilVDR: A Compact End-to-End Visual Document Retriever via Dual-Student Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.CV"],"primary_cat":"cs.IR","authors_text":"Yao Zhang, Yu Xiao, Zhuchenyang Liu, Ziyi Wang","submitted_at":"2026-08-11T08:23:39Z","abstract_excerpt":"Visual document retrieval (VDR) is dominated by multi-billion-parameter models that are slow to index at full corpus scale and expensive to serve. Prior compression routes either train a smaller multi-vector encoder from scratch or distil only the query side; neither yields a compact single-vector retriever end-to-end. We present DistilVDR, a 524M end-to-end VDR system distilled bilaterally from a single 8B vision-language teacher under a pointwise cosine alignment loss. All supervision comes from the frozen teacher's embedding space, which was itself trained with relevance supervision, so the"},"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":"2608.10636","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-08-11T08:23:39Z","cross_cats_sorted":["cs.CL","cs.CV"],"title_canon_sha256":"1977f08ef5ecdb04403a9ea69949601c1a4973036275da2149f00670bc2ca913","abstract_canon_sha256":"c9eb57ca97f877cb00a2234fe9d48fd7a5e5b8a3a4b0d82051273ae668e27dc5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-12T01:22:55.239369Z","signature_b64":"PFYBpRbrtLN7y7w0dUJUlOwq/X+9HvSL8uYE9eebbFF2JTwf18DHAexQaTS33ltex3EGA5KfU43X7qUc5KnxBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e79c8c8e987538c0d2115b8c7f51a8605efea5c415f23b924d6bce11663b5675","last_reissued_at":"2026-08-12T01:22:55.237609Z","signature_status":"signed_v1","first_computed_at":"2026-08-12T01:22:55.237609Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DistilVDR: A Compact End-to-End Visual Document Retriever via Dual-Student Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.CV"],"primary_cat":"cs.IR","authors_text":"Yao Zhang, Yu Xiao, Zhuchenyang Liu, Ziyi Wang","submitted_at":"2026-08-11T08:23:39Z","abstract_excerpt":"Visual document retrieval (VDR) is dominated by multi-billion-parameter models that are slow to index at full corpus scale and expensive to serve. Prior compression routes either train a smaller multi-vector encoder from scratch or distil only the query side; neither yields a compact single-vector retriever end-to-end. We present DistilVDR, a 524M end-to-end VDR system distilled bilaterally from a single 8B vision-language teacher under a pointwise cosine alignment loss. All supervision comes from the frozen teacher's embedding space, which was itself trained with relevance supervision, so the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.10636","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/2608.10636/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":"2608.10636","created_at":"2026-08-12T01:22:55.241436+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.10636v1","created_at":"2026-08-12T01:22:55.241436+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.10636","created_at":"2026-08-12T01:22:55.241436+00:00"},{"alias_kind":"pith_short_12","alias_value":"46OIZDUYOU4M","created_at":"2026-08-12T01:22:55.241436+00:00"},{"alias_kind":"pith_short_16","alias_value":"46OIZDUYOU4MBUQR","created_at":"2026-08-12T01:22:55.241436+00:00"},{"alias_kind":"pith_short_8","alias_value":"46OIZDUY","created_at":"2026-08-12T01:22:55.241436+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/46OIZDUYOU4MBUQRLOGH6UNIMB","json":"https://pith.science/pith/46OIZDUYOU4MBUQRLOGH6UNIMB.json","graph_json":"https://pith.science/api/pith-number/46OIZDUYOU4MBUQRLOGH6UNIMB/graph.json","events_json":"https://pith.science/api/pith-number/46OIZDUYOU4MBUQRLOGH6UNIMB/events.json","paper":"https://pith.science/paper/46OIZDUY"},"agent_actions":{"view_html":"https://pith.science/pith/46OIZDUYOU4MBUQRLOGH6UNIMB","download_json":"https://pith.science/pith/46OIZDUYOU4MBUQRLOGH6UNIMB.json","view_paper":"https://pith.science/paper/46OIZDUY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.10636&json=true","fetch_graph":"https://pith.science/api/pith-number/46OIZDUYOU4MBUQRLOGH6UNIMB/graph.json","fetch_events":"https://pith.science/api/pith-number/46OIZDUYOU4MBUQRLOGH6UNIMB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/46OIZDUYOU4MBUQRLOGH6UNIMB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/46OIZDUYOU4MBUQRLOGH6UNIMB/action/storage_attestation","attest_author":"https://pith.science/pith/46OIZDUYOU4MBUQRLOGH6UNIMB/action/author_attestation","sign_citation":"https://pith.science/pith/46OIZDUYOU4MBUQRLOGH6UNIMB/action/citation_signature","submit_replication":"https://pith.science/pith/46OIZDUYOU4MBUQRLOGH6UNIMB/action/replication_record"}},"created_at":"2026-08-12T01:22:55.241436+00:00","updated_at":"2026-08-12T01:22:55.241436+00:00"}