{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ZMVJR55C6LCUZTHYFI6L3ZVT5B","short_pith_number":"pith:ZMVJR55C","schema_version":"1.0","canonical_sha256":"cb2a98f7a2f2c54cccf82a3cbde6b3e87a28ce71c0e50a6e0dda973838980277","source":{"kind":"arxiv","id":"2504.09816","version":1},"attestation_state":"computed","paper":{"title":"Augmented Relevance Datasets with Fine-Tuned Small LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.IR","authors_text":"Matyas Amrouche, Quentin Fitte-Rey, Romain Deveaud","submitted_at":"2025-04-14T02:35:00Z","abstract_excerpt":"Building high-quality datasets and labeling query-document relevance are essential yet resource-intensive tasks, requiring detailed guidelines and substantial effort from human annotators. This paper explores the use of small, fine-tuned large language models (LLMs) to automate relevance assessment, with a focus on improving ranking models' performance by augmenting their training dataset. We fine-tuned small LLMs to enhance relevance assessments, thereby improving dataset creation quality for downstream ranking model training. Our experiments demonstrate that these fine-tuned small LLMs not o"},"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":"2504.09816","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-04-14T02:35:00Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"ecac35758c5c69bac03f351eacabc872da17ea2fb4863966d1d0daafec411c10","abstract_canon_sha256":"812608a967318fba701c97a990aa89f479b644a9fd40a6c738cd2bbda3959b75"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:48:43.135221Z","signature_b64":"jEXj1F/4VKz1ghuJEBOXl7opvhAIWA9KmERb7cBjBc3RLE6KLDnVrv+kN5bP4Laj12hA/hIBqZygMcCSM/1PDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cb2a98f7a2f2c54cccf82a3cbde6b3e87a28ce71c0e50a6e0dda973838980277","last_reissued_at":"2026-07-05T10:48:43.134796Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:48:43.134796Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Augmented Relevance Datasets with Fine-Tuned Small LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.IR","authors_text":"Matyas Amrouche, Quentin Fitte-Rey, Romain Deveaud","submitted_at":"2025-04-14T02:35:00Z","abstract_excerpt":"Building high-quality datasets and labeling query-document relevance are essential yet resource-intensive tasks, requiring detailed guidelines and substantial effort from human annotators. This paper explores the use of small, fine-tuned large language models (LLMs) to automate relevance assessment, with a focus on improving ranking models' performance by augmenting their training dataset. We fine-tuned small LLMs to enhance relevance assessments, thereby improving dataset creation quality for downstream ranking model training. Our experiments demonstrate that these fine-tuned small LLMs not o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.09816","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/2504.09816/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":"2504.09816","created_at":"2026-07-05T10:48:43.134852+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.09816v1","created_at":"2026-07-05T10:48:43.134852+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.09816","created_at":"2026-07-05T10:48:43.134852+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZMVJR55C6LCU","created_at":"2026-07-05T10:48:43.134852+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZMVJR55C6LCUZTHY","created_at":"2026-07-05T10:48:43.134852+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZMVJR55C","created_at":"2026-07-05T10:48:43.134852+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.00785","citing_title":"HIERA: Hierarchical Multi-Agent Relevance Assessment for Content Discovery Systems","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZMVJR55C6LCUZTHYFI6L3ZVT5B","json":"https://pith.science/pith/ZMVJR55C6LCUZTHYFI6L3ZVT5B.json","graph_json":"https://pith.science/api/pith-number/ZMVJR55C6LCUZTHYFI6L3ZVT5B/graph.json","events_json":"https://pith.science/api/pith-number/ZMVJR55C6LCUZTHYFI6L3ZVT5B/events.json","paper":"https://pith.science/paper/ZMVJR55C"},"agent_actions":{"view_html":"https://pith.science/pith/ZMVJR55C6LCUZTHYFI6L3ZVT5B","download_json":"https://pith.science/pith/ZMVJR55C6LCUZTHYFI6L3ZVT5B.json","view_paper":"https://pith.science/paper/ZMVJR55C","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.09816&json=true","fetch_graph":"https://pith.science/api/pith-number/ZMVJR55C6LCUZTHYFI6L3ZVT5B/graph.json","fetch_events":"https://pith.science/api/pith-number/ZMVJR55C6LCUZTHYFI6L3ZVT5B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZMVJR55C6LCUZTHYFI6L3ZVT5B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZMVJR55C6LCUZTHYFI6L3ZVT5B/action/storage_attestation","attest_author":"https://pith.science/pith/ZMVJR55C6LCUZTHYFI6L3ZVT5B/action/author_attestation","sign_citation":"https://pith.science/pith/ZMVJR55C6LCUZTHYFI6L3ZVT5B/action/citation_signature","submit_replication":"https://pith.science/pith/ZMVJR55C6LCUZTHYFI6L3ZVT5B/action/replication_record"}},"created_at":"2026-07-05T10:48:43.134852+00:00","updated_at":"2026-07-05T10:48:43.134852+00:00"}