{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:PULEPN2MAYQIQVY6AD7BWYA5HQ","short_pith_number":"pith:PULEPN2M","schema_version":"1.0","canonical_sha256":"7d1647b74c062088571e00fe1b601d3c2ec00bd1d08885599ed4f4045ee57087","source":{"kind":"arxiv","id":"2210.03809","version":2},"attestation_state":"computed","paper":{"title":"Retrieval Augmented Visual Question Answering with Outside Knowledge","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bill Byrne, Weizhe Lin","submitted_at":"2022-10-07T20:35:58Z","abstract_excerpt":"Outside-Knowledge Visual Question Answering (OK-VQA) is a challenging VQA task that requires retrieval of external knowledge to answer questions about images. Recent OK-VQA systems use Dense Passage Retrieval (DPR) to retrieve documents from external knowledge bases, such as Wikipedia, but with DPR trained separately from answer generation, introducing a potential limit on the overall system performance. Instead, we propose a joint training scheme which includes differentiable DPR integrated with answer generation so that the system can be trained in an end-to-end fashion. Our experiments show"},"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":"2210.03809","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-10-07T20:35:58Z","cross_cats_sorted":[],"title_canon_sha256":"8d736974eb0b10269c596ea8a99764ca976931692598bf1c7e26621d3b627375","abstract_canon_sha256":"41401d520672548adb44f70105270009d8248f9278e16839f71562808ce21572"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:11:43.394552Z","signature_b64":"4AP1ONugNcRU297kL/3K9CzN/PdcgDoWsFVoH4AQmR55H2JTRAE038+qK3t515HGJoGYIUQaHPV8YCli0I/dBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7d1647b74c062088571e00fe1b601d3c2ec00bd1d08885599ed4f4045ee57087","last_reissued_at":"2026-07-05T05:11:43.394064Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:11:43.394064Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Retrieval Augmented Visual Question Answering with Outside Knowledge","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bill Byrne, Weizhe Lin","submitted_at":"2022-10-07T20:35:58Z","abstract_excerpt":"Outside-Knowledge Visual Question Answering (OK-VQA) is a challenging VQA task that requires retrieval of external knowledge to answer questions about images. Recent OK-VQA systems use Dense Passage Retrieval (DPR) to retrieve documents from external knowledge bases, such as Wikipedia, but with DPR trained separately from answer generation, introducing a potential limit on the overall system performance. Instead, we propose a joint training scheme which includes differentiable DPR integrated with answer generation so that the system can be trained in an end-to-end fashion. Our experiments show"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.03809","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/2210.03809/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":"2210.03809","created_at":"2026-07-05T05:11:43.394125+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.03809v2","created_at":"2026-07-05T05:11:43.394125+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.03809","created_at":"2026-07-05T05:11:43.394125+00:00"},{"alias_kind":"pith_short_12","alias_value":"PULEPN2MAYQI","created_at":"2026-07-05T05:11:43.394125+00:00"},{"alias_kind":"pith_short_16","alias_value":"PULEPN2MAYQIQVY6","created_at":"2026-07-05T05:11:43.394125+00:00"},{"alias_kind":"pith_short_8","alias_value":"PULEPN2M","created_at":"2026-07-05T05:11:43.394125+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2303.16199","citing_title":"LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention","ref_index":139,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PULEPN2MAYQIQVY6AD7BWYA5HQ","json":"https://pith.science/pith/PULEPN2MAYQIQVY6AD7BWYA5HQ.json","graph_json":"https://pith.science/api/pith-number/PULEPN2MAYQIQVY6AD7BWYA5HQ/graph.json","events_json":"https://pith.science/api/pith-number/PULEPN2MAYQIQVY6AD7BWYA5HQ/events.json","paper":"https://pith.science/paper/PULEPN2M"},"agent_actions":{"view_html":"https://pith.science/pith/PULEPN2MAYQIQVY6AD7BWYA5HQ","download_json":"https://pith.science/pith/PULEPN2MAYQIQVY6AD7BWYA5HQ.json","view_paper":"https://pith.science/paper/PULEPN2M","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.03809&json=true","fetch_graph":"https://pith.science/api/pith-number/PULEPN2MAYQIQVY6AD7BWYA5HQ/graph.json","fetch_events":"https://pith.science/api/pith-number/PULEPN2MAYQIQVY6AD7BWYA5HQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PULEPN2MAYQIQVY6AD7BWYA5HQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PULEPN2MAYQIQVY6AD7BWYA5HQ/action/storage_attestation","attest_author":"https://pith.science/pith/PULEPN2MAYQIQVY6AD7BWYA5HQ/action/author_attestation","sign_citation":"https://pith.science/pith/PULEPN2MAYQIQVY6AD7BWYA5HQ/action/citation_signature","submit_replication":"https://pith.science/pith/PULEPN2MAYQIQVY6AD7BWYA5HQ/action/replication_record"}},"created_at":"2026-07-05T05:11:43.394125+00:00","updated_at":"2026-07-05T05:11:43.394125+00:00"}