{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:J3NWZQI2VX2D34MIJPMMX2EYD7","short_pith_number":"pith:J3NWZQI2","schema_version":"1.0","canonical_sha256":"4edb6cc11aadf43df1884bd8cbe8981fec45ac54f85f9b66366d43cd80a4ef1e","source":{"kind":"arxiv","id":"2411.06710","version":2},"attestation_state":"computed","paper":{"title":"Model Fusion through Bayesian Optimization in Language Model Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Chaeyun Jang, Hyungi Lee, Juho Lee, Jungtaek Kim","submitted_at":"2024-11-11T04:36:58Z","abstract_excerpt":"Fine-tuning pre-trained models for downstream tasks is a widely adopted technique known for its adaptability and reliability across various domains. Despite its conceptual simplicity, fine-tuning entails several troublesome engineering choices, such as selecting hyperparameters and determining checkpoints from an optimization trajectory. To tackle the difficulty of choosing the best model, one effective solution is model fusion, which combines multiple models in a parameter space. However, we observe a large discrepancy between loss and metric landscapes during the fine-tuning of pre-trained l"},"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":"2411.06710","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-11-11T04:36:58Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"7120419f8b1584496d9df84fb37201210fa66fb849acae5d3163b8fe8a265dd1","abstract_canon_sha256":"459a5717ddb8208abe3b9b7e1d142da500b5f729a3f70572d94cb9ba91dbf6dc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:54:29.560667Z","signature_b64":"GhVXfvimCg1/1XeSKftemFqNi0ltlK5a2hwvPgXl2Vu5wN+VWn9PXlmS3JdCARrofTSxwl6vjvzx5ZgHYM/JAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4edb6cc11aadf43df1884bd8cbe8981fec45ac54f85f9b66366d43cd80a4ef1e","last_reissued_at":"2026-07-05T09:54:29.560185Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:54:29.560185Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Model Fusion through Bayesian Optimization in Language Model Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Chaeyun Jang, Hyungi Lee, Juho Lee, Jungtaek Kim","submitted_at":"2024-11-11T04:36:58Z","abstract_excerpt":"Fine-tuning pre-trained models for downstream tasks is a widely adopted technique known for its adaptability and reliability across various domains. Despite its conceptual simplicity, fine-tuning entails several troublesome engineering choices, such as selecting hyperparameters and determining checkpoints from an optimization trajectory. To tackle the difficulty of choosing the best model, one effective solution is model fusion, which combines multiple models in a parameter space. However, we observe a large discrepancy between loss and metric landscapes during the fine-tuning of pre-trained l"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.06710","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/2411.06710/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":"2411.06710","created_at":"2026-07-05T09:54:29.560243+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.06710v2","created_at":"2026-07-05T09:54:29.560243+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.06710","created_at":"2026-07-05T09:54:29.560243+00:00"},{"alias_kind":"pith_short_12","alias_value":"J3NWZQI2VX2D","created_at":"2026-07-05T09:54:29.560243+00:00"},{"alias_kind":"pith_short_16","alias_value":"J3NWZQI2VX2D34MI","created_at":"2026-07-05T09:54:29.560243+00:00"},{"alias_kind":"pith_short_8","alias_value":"J3NWZQI2","created_at":"2026-07-05T09:54:29.560243+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.17151","citing_title":"Bayesian Optimization for Enhanced Language Models: Optimizing Acquisition Functions","ref_index":1,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/J3NWZQI2VX2D34MIJPMMX2EYD7","json":"https://pith.science/pith/J3NWZQI2VX2D34MIJPMMX2EYD7.json","graph_json":"https://pith.science/api/pith-number/J3NWZQI2VX2D34MIJPMMX2EYD7/graph.json","events_json":"https://pith.science/api/pith-number/J3NWZQI2VX2D34MIJPMMX2EYD7/events.json","paper":"https://pith.science/paper/J3NWZQI2"},"agent_actions":{"view_html":"https://pith.science/pith/J3NWZQI2VX2D34MIJPMMX2EYD7","download_json":"https://pith.science/pith/J3NWZQI2VX2D34MIJPMMX2EYD7.json","view_paper":"https://pith.science/paper/J3NWZQI2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.06710&json=true","fetch_graph":"https://pith.science/api/pith-number/J3NWZQI2VX2D34MIJPMMX2EYD7/graph.json","fetch_events":"https://pith.science/api/pith-number/J3NWZQI2VX2D34MIJPMMX2EYD7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/J3NWZQI2VX2D34MIJPMMX2EYD7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/J3NWZQI2VX2D34MIJPMMX2EYD7/action/storage_attestation","attest_author":"https://pith.science/pith/J3NWZQI2VX2D34MIJPMMX2EYD7/action/author_attestation","sign_citation":"https://pith.science/pith/J3NWZQI2VX2D34MIJPMMX2EYD7/action/citation_signature","submit_replication":"https://pith.science/pith/J3NWZQI2VX2D34MIJPMMX2EYD7/action/replication_record"}},"created_at":"2026-07-05T09:54:29.560243+00:00","updated_at":"2026-07-05T09:54:29.560243+00:00"}