{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:G4G4F4XFH7U6DHEFE44MYNFUBJ","short_pith_number":"pith:G4G4F4XF","schema_version":"1.0","canonical_sha256":"370dc2f2e53fe9e19c852738cc34b40a50a7fbcdfe1b2465759d26477e956476","source":{"kind":"arxiv","id":"2112.07577","version":3},"attestation_state":"computed","paper":{"title":"GPL: Generative Pseudo Labeling for Unsupervised Domain Adaptation of Dense Retrieval","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CL","authors_text":"Iryna Gurevych, Kexin Wang, Nandan Thakur, Nils Reimers","submitted_at":"2021-12-14T17:34:43Z","abstract_excerpt":"Dense retrieval approaches can overcome the lexical gap and lead to significantly improved search results. However, they require large amounts of training data which is not available for most domains. As shown in previous work (Thakur et al., 2021b), the performance of dense retrievers severely degrades under a domain shift. This limits the usage of dense retrieval approaches to only a few domains with large training datasets.\n  In this paper, we propose the novel unsupervised domain adaptation method Generative Pseudo Labeling (GPL), which combines a query generator with pseudo labeling from "},"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":"2112.07577","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2021-12-14T17:34:43Z","cross_cats_sorted":["cs.IR"],"title_canon_sha256":"7c6a983bfb98a58d1b2f5f9b71b249fdc38a4f6b7b967dfea1db97295313e06d","abstract_canon_sha256":"edd2f7ef3ebe1f39dfbe53dcc495bce5a1cd5ec6e1cbca719e6c03294c7f7074"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:17:19.417410Z","signature_b64":"5e/jlgQJXUJUAdh3PUnYreeXPEdlQH0SscMZV+i6IPkNUiEF1bO6qChU5n5NeEyVaEQAdqaDPb0H7FSV4qQNAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"370dc2f2e53fe9e19c852738cc34b40a50a7fbcdfe1b2465759d26477e956476","last_reissued_at":"2026-07-05T04:17:19.416956Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:17:19.416956Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GPL: Generative Pseudo Labeling for Unsupervised Domain Adaptation of Dense Retrieval","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CL","authors_text":"Iryna Gurevych, Kexin Wang, Nandan Thakur, Nils Reimers","submitted_at":"2021-12-14T17:34:43Z","abstract_excerpt":"Dense retrieval approaches can overcome the lexical gap and lead to significantly improved search results. However, they require large amounts of training data which is not available for most domains. As shown in previous work (Thakur et al., 2021b), the performance of dense retrievers severely degrades under a domain shift. This limits the usage of dense retrieval approaches to only a few domains with large training datasets.\n  In this paper, we propose the novel unsupervised domain adaptation method Generative Pseudo Labeling (GPL), which combines a query generator with pseudo labeling from "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2112.07577","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/2112.07577/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":"2112.07577","created_at":"2026-07-05T04:17:19.417003+00:00"},{"alias_kind":"arxiv_version","alias_value":"2112.07577v3","created_at":"2026-07-05T04:17:19.417003+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2112.07577","created_at":"2026-07-05T04:17:19.417003+00:00"},{"alias_kind":"pith_short_12","alias_value":"G4G4F4XFH7U6","created_at":"2026-07-05T04:17:19.417003+00:00"},{"alias_kind":"pith_short_16","alias_value":"G4G4F4XFH7U6DHEF","created_at":"2026-07-05T04:17:19.417003+00:00"},{"alias_kind":"pith_short_8","alias_value":"G4G4F4XF","created_at":"2026-07-05T04:17:19.417003+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.27291","citing_title":"Designing Reward Signals for Portable Query Generation: A Case Study in Industrial Semantic Job Search","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2504.21015","citing_title":"Don't Retrieve, Generate: Prompting LLMs for Synthetic Training Data in Dense Retrieval","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25142","citing_title":"UnIte: Uncertainty-based Iterative Document Sampling for Domain Adaptation in Information Retrieval","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/G4G4F4XFH7U6DHEFE44MYNFUBJ","json":"https://pith.science/pith/G4G4F4XFH7U6DHEFE44MYNFUBJ.json","graph_json":"https://pith.science/api/pith-number/G4G4F4XFH7U6DHEFE44MYNFUBJ/graph.json","events_json":"https://pith.science/api/pith-number/G4G4F4XFH7U6DHEFE44MYNFUBJ/events.json","paper":"https://pith.science/paper/G4G4F4XF"},"agent_actions":{"view_html":"https://pith.science/pith/G4G4F4XFH7U6DHEFE44MYNFUBJ","download_json":"https://pith.science/pith/G4G4F4XFH7U6DHEFE44MYNFUBJ.json","view_paper":"https://pith.science/paper/G4G4F4XF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2112.07577&json=true","fetch_graph":"https://pith.science/api/pith-number/G4G4F4XFH7U6DHEFE44MYNFUBJ/graph.json","fetch_events":"https://pith.science/api/pith-number/G4G4F4XFH7U6DHEFE44MYNFUBJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G4G4F4XFH7U6DHEFE44MYNFUBJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G4G4F4XFH7U6DHEFE44MYNFUBJ/action/storage_attestation","attest_author":"https://pith.science/pith/G4G4F4XFH7U6DHEFE44MYNFUBJ/action/author_attestation","sign_citation":"https://pith.science/pith/G4G4F4XFH7U6DHEFE44MYNFUBJ/action/citation_signature","submit_replication":"https://pith.science/pith/G4G4F4XFH7U6DHEFE44MYNFUBJ/action/replication_record"}},"created_at":"2026-07-05T04:17:19.417003+00:00","updated_at":"2026-07-05T04:17:19.417003+00:00"}