{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:EYKK2WKNXPH5WXWCJBZF6VOKXD","short_pith_number":"pith:EYKK2WKN","schema_version":"1.0","canonical_sha256":"2614ad594dbbcfdb5ec248725f55cab8fe3bed430e109f60955b2f03f865f954","source":{"kind":"arxiv","id":"2403.13187","version":2},"attestation_state":"computed","paper":{"title":"Evolutionary Optimization of Model Merging Recipes","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.NE","authors_text":"David Ha, Makoto Shing, Qi Sun, Takuya Akiba, Yujin Tang","submitted_at":"2024-03-19T22:56:53Z","abstract_excerpt":"Large language models (LLMs) have become increasingly capable, but their development often requires substantial computational resources. While model merging has emerged as a cost-effective promising approach for creating new models by combining existing ones, it currently relies on human intuition and domain knowledge, limiting its potential. Here, we propose an evolutionary approach that overcomes this limitation by automatically discovering effective combinations of diverse open-source models, harnessing their collective intelligence without requiring extensive additional training data or co"},"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":"2403.13187","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2024-03-19T22:56:53Z","cross_cats_sorted":[],"title_canon_sha256":"250da58d2cd9c1c5f12d422f787529f3e69c7a9a7dbc6bee27ed1d250aaede11","abstract_canon_sha256":"db1242bed8c7f5a903c48b07b41d3d94ec1ab0b7fa335e2de73170c54ca0273c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:05:42.105000Z","signature_b64":"KlQNoRDyuIBogGHL79HJHy5H188Ik1FEQJEW3pPT6bFfyq4hQsW8gBbYImQ+HIFm5kUNAY6mtOPUlatTMAFBDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2614ad594dbbcfdb5ec248725f55cab8fe3bed430e109f60955b2f03f865f954","last_reissued_at":"2026-07-05T10:05:42.104542Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:05:42.104542Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evolutionary Optimization of Model Merging Recipes","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.NE","authors_text":"David Ha, Makoto Shing, Qi Sun, Takuya Akiba, Yujin Tang","submitted_at":"2024-03-19T22:56:53Z","abstract_excerpt":"Large language models (LLMs) have become increasingly capable, but their development often requires substantial computational resources. While model merging has emerged as a cost-effective promising approach for creating new models by combining existing ones, it currently relies on human intuition and domain knowledge, limiting its potential. Here, we propose an evolutionary approach that overcomes this limitation by automatically discovering effective combinations of diverse open-source models, harnessing their collective intelligence without requiring extensive additional training data or co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.13187","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/2403.13187/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":"2403.13187","created_at":"2026-07-05T10:05:42.104614+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.13187v2","created_at":"2026-07-05T10:05:42.104614+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.13187","created_at":"2026-07-05T10:05:42.104614+00:00"},{"alias_kind":"pith_short_12","alias_value":"EYKK2WKNXPH5","created_at":"2026-07-05T10:05:42.104614+00:00"},{"alias_kind":"pith_short_16","alias_value":"EYKK2WKNXPH5WXWC","created_at":"2026-07-05T10:05:42.104614+00:00"},{"alias_kind":"pith_short_8","alias_value":"EYKK2WKN","created_at":"2026-07-05T10:05:42.104614+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.28649","citing_title":"Interpretability-Guided Layer Selection over Subspace Projection: SAEs as Stethoscopes, Not Scalpels, for Raw Task Vector Model Editing","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2604.01674","citing_title":"Can Heterogeneous Language Models Be Fused?","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20803","citing_title":"Tunable MAGMAX: Preference-Aware Model Merging for Continual Learning","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14386","citing_title":"Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model Reasoning","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12326","citing_title":"Black-Box Optimization of Mixed Binary-Continuous Variables: Challenges and Opportunities in Evolutionary Model Merging","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26211","citing_title":"OMEGA: Optimizing Machine Learning by Evaluating Generated Algorithms","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EYKK2WKNXPH5WXWCJBZF6VOKXD","json":"https://pith.science/pith/EYKK2WKNXPH5WXWCJBZF6VOKXD.json","graph_json":"https://pith.science/api/pith-number/EYKK2WKNXPH5WXWCJBZF6VOKXD/graph.json","events_json":"https://pith.science/api/pith-number/EYKK2WKNXPH5WXWCJBZF6VOKXD/events.json","paper":"https://pith.science/paper/EYKK2WKN"},"agent_actions":{"view_html":"https://pith.science/pith/EYKK2WKNXPH5WXWCJBZF6VOKXD","download_json":"https://pith.science/pith/EYKK2WKNXPH5WXWCJBZF6VOKXD.json","view_paper":"https://pith.science/paper/EYKK2WKN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.13187&json=true","fetch_graph":"https://pith.science/api/pith-number/EYKK2WKNXPH5WXWCJBZF6VOKXD/graph.json","fetch_events":"https://pith.science/api/pith-number/EYKK2WKNXPH5WXWCJBZF6VOKXD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EYKK2WKNXPH5WXWCJBZF6VOKXD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EYKK2WKNXPH5WXWCJBZF6VOKXD/action/storage_attestation","attest_author":"https://pith.science/pith/EYKK2WKNXPH5WXWCJBZF6VOKXD/action/author_attestation","sign_citation":"https://pith.science/pith/EYKK2WKNXPH5WXWCJBZF6VOKXD/action/citation_signature","submit_replication":"https://pith.science/pith/EYKK2WKNXPH5WXWCJBZF6VOKXD/action/replication_record"}},"created_at":"2026-07-05T10:05:42.104614+00:00","updated_at":"2026-07-05T10:05:42.104614+00:00"}