{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HHSF3FPMG3IO6OA5YHZZN3CDJN","short_pith_number":"pith:HHSF3FPM","schema_version":"1.0","canonical_sha256":"39e45d95ec36d0ef381dc1f396ec434b6a717dfd51fcd46b550dc18089f06b0e","source":{"kind":"arxiv","id":"2505.07166","version":1},"attestation_state":"computed","paper":{"title":"Pre-training vs. Fine-tuning: A Reproducibility Study on Dense Retrieval Knowledge Acquisition","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.IR","authors_text":"Guido Zuccon, Shuai Wang, Zheng Yao","submitted_at":"2025-05-12T01:24:00Z","abstract_excerpt":"Dense retrievers utilize pre-trained backbone language models (e.g., BERT, LLaMA) that are fine-tuned via contrastive learning to perform the task of encoding text into sense representations that can be then compared via a shallow similarity operation, e.g. inner product. Recent research has questioned the role of fine-tuning vs. that of pre-training within dense retrievers, specifically arguing that retrieval knowledge is primarily gained during pre-training, meaning knowledge not acquired during pre-training cannot be sub-sequentially acquired via fine-tuning. We revisit this idea here as th"},"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":"2505.07166","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-05-12T01:24:00Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"324675eba131d97d2ceba27acfed34b70820064005b71fcd3ee68e7cb2935155","abstract_canon_sha256":"f1eb373a830dbabfa289a335eec27a0ba180a7194d1d0ad060fd171d654cffbc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:01:43.546393Z","signature_b64":"tU9Bp9PfSyy2ueJKE73nR/uU3uBbjJRBfrDokzWkM0pfZ3em87GtCXUQ3DyeSpdRDCzRu/yGqd9UGaPkAI4CAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"39e45d95ec36d0ef381dc1f396ec434b6a717dfd51fcd46b550dc18089f06b0e","last_reissued_at":"2026-07-05T11:01:43.545737Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:01:43.545737Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Pre-training vs. Fine-tuning: A Reproducibility Study on Dense Retrieval Knowledge Acquisition","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.IR","authors_text":"Guido Zuccon, Shuai Wang, Zheng Yao","submitted_at":"2025-05-12T01:24:00Z","abstract_excerpt":"Dense retrievers utilize pre-trained backbone language models (e.g., BERT, LLaMA) that are fine-tuned via contrastive learning to perform the task of encoding text into sense representations that can be then compared via a shallow similarity operation, e.g. inner product. Recent research has questioned the role of fine-tuning vs. that of pre-training within dense retrievers, specifically arguing that retrieval knowledge is primarily gained during pre-training, meaning knowledge not acquired during pre-training cannot be sub-sequentially acquired via fine-tuning. We revisit this idea here as th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.07166","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/2505.07166/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":"2505.07166","created_at":"2026-07-05T11:01:43.545790+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.07166v1","created_at":"2026-07-05T11:01:43.545790+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.07166","created_at":"2026-07-05T11:01:43.545790+00:00"},{"alias_kind":"pith_short_12","alias_value":"HHSF3FPMG3IO","created_at":"2026-07-05T11:01:43.545790+00:00"},{"alias_kind":"pith_short_16","alias_value":"HHSF3FPMG3IO6OA5","created_at":"2026-07-05T11:01:43.545790+00:00"},{"alias_kind":"pith_short_8","alias_value":"HHSF3FPM","created_at":"2026-07-05T11:01:43.545790+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/HHSF3FPMG3IO6OA5YHZZN3CDJN","json":"https://pith.science/pith/HHSF3FPMG3IO6OA5YHZZN3CDJN.json","graph_json":"https://pith.science/api/pith-number/HHSF3FPMG3IO6OA5YHZZN3CDJN/graph.json","events_json":"https://pith.science/api/pith-number/HHSF3FPMG3IO6OA5YHZZN3CDJN/events.json","paper":"https://pith.science/paper/HHSF3FPM"},"agent_actions":{"view_html":"https://pith.science/pith/HHSF3FPMG3IO6OA5YHZZN3CDJN","download_json":"https://pith.science/pith/HHSF3FPMG3IO6OA5YHZZN3CDJN.json","view_paper":"https://pith.science/paper/HHSF3FPM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.07166&json=true","fetch_graph":"https://pith.science/api/pith-number/HHSF3FPMG3IO6OA5YHZZN3CDJN/graph.json","fetch_events":"https://pith.science/api/pith-number/HHSF3FPMG3IO6OA5YHZZN3CDJN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HHSF3FPMG3IO6OA5YHZZN3CDJN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HHSF3FPMG3IO6OA5YHZZN3CDJN/action/storage_attestation","attest_author":"https://pith.science/pith/HHSF3FPMG3IO6OA5YHZZN3CDJN/action/author_attestation","sign_citation":"https://pith.science/pith/HHSF3FPMG3IO6OA5YHZZN3CDJN/action/citation_signature","submit_replication":"https://pith.science/pith/HHSF3FPMG3IO6OA5YHZZN3CDJN/action/replication_record"}},"created_at":"2026-07-05T11:01:43.545790+00:00","updated_at":"2026-07-05T11:01:43.545790+00:00"}