{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:OTK53WRS5WWGEKM6G3PHYVUUWF","short_pith_number":"pith:OTK53WRS","schema_version":"1.0","canonical_sha256":"74d5ddda32edac62299e36de7c5694b179c3a3b3188ed232ad4f8aa3d1c64ed0","source":{"kind":"arxiv","id":"2502.12706","version":2},"attestation_state":"computed","paper":{"title":"Scalable Model Merging with Progressive Layer-wise Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jiazheng Li, Jing Xu, Jingzhao Zhang","submitted_at":"2025-02-18T10:15:18Z","abstract_excerpt":"Model merging offers an effective way to integrate the capabilities of multiple fine-tuned models. However, the performance degradation of the merged model remains a challenge, particularly when none or few data are available. This paper first highlights the necessity of domain-specific data for model merging by proving that data-agnostic algorithms can have arbitrarily bad worst-case performance. Building on this theoretical insight, we explore the relationship between model merging and distillation, introducing a novel few-shot merging algorithm, ProDistill (Progressive Layer-wise Distillati"},"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":"2502.12706","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-18T10:15:18Z","cross_cats_sorted":[],"title_canon_sha256":"e0e150dd6c04d55dd555c5b07f750745c6b0e6b46e315803cadc39108ec07b8f","abstract_canon_sha256":"fcc8bef8ac70d33c51a810af6944813ed53cb88cc1e6a1d1f49593271c6b8205"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:10:05.112161Z","signature_b64":"VOs8JLRQl0CZ0i47ViXZKm0FKLObBQUE9dUsb1LNgFp3BgAM8UQ4+n+89PiuwOw7EobjrY8wGRuMqJc1mwu5Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"74d5ddda32edac62299e36de7c5694b179c3a3b3188ed232ad4f8aa3d1c64ed0","last_reissued_at":"2026-07-05T11:10:05.111651Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:10:05.111651Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scalable Model Merging with Progressive Layer-wise Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jiazheng Li, Jing Xu, Jingzhao Zhang","submitted_at":"2025-02-18T10:15:18Z","abstract_excerpt":"Model merging offers an effective way to integrate the capabilities of multiple fine-tuned models. However, the performance degradation of the merged model remains a challenge, particularly when none or few data are available. This paper first highlights the necessity of domain-specific data for model merging by proving that data-agnostic algorithms can have arbitrarily bad worst-case performance. Building on this theoretical insight, we explore the relationship between model merging and distillation, introducing a novel few-shot merging algorithm, ProDistill (Progressive Layer-wise Distillati"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.12706","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/2502.12706/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":"2502.12706","created_at":"2026-07-05T11:10:05.111711+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.12706v2","created_at":"2026-07-05T11:10:05.111711+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.12706","created_at":"2026-07-05T11:10:05.111711+00:00"},{"alias_kind":"pith_short_12","alias_value":"OTK53WRS5WWG","created_at":"2026-07-05T11:10:05.111711+00:00"},{"alias_kind":"pith_short_16","alias_value":"OTK53WRS5WWGEKM6","created_at":"2026-07-05T11:10:05.111711+00:00"},{"alias_kind":"pith_short_8","alias_value":"OTK53WRS","created_at":"2026-07-05T11:10:05.111711+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/OTK53WRS5WWGEKM6G3PHYVUUWF","json":"https://pith.science/pith/OTK53WRS5WWGEKM6G3PHYVUUWF.json","graph_json":"https://pith.science/api/pith-number/OTK53WRS5WWGEKM6G3PHYVUUWF/graph.json","events_json":"https://pith.science/api/pith-number/OTK53WRS5WWGEKM6G3PHYVUUWF/events.json","paper":"https://pith.science/paper/OTK53WRS"},"agent_actions":{"view_html":"https://pith.science/pith/OTK53WRS5WWGEKM6G3PHYVUUWF","download_json":"https://pith.science/pith/OTK53WRS5WWGEKM6G3PHYVUUWF.json","view_paper":"https://pith.science/paper/OTK53WRS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.12706&json=true","fetch_graph":"https://pith.science/api/pith-number/OTK53WRS5WWGEKM6G3PHYVUUWF/graph.json","fetch_events":"https://pith.science/api/pith-number/OTK53WRS5WWGEKM6G3PHYVUUWF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OTK53WRS5WWGEKM6G3PHYVUUWF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OTK53WRS5WWGEKM6G3PHYVUUWF/action/storage_attestation","attest_author":"https://pith.science/pith/OTK53WRS5WWGEKM6G3PHYVUUWF/action/author_attestation","sign_citation":"https://pith.science/pith/OTK53WRS5WWGEKM6G3PHYVUUWF/action/citation_signature","submit_replication":"https://pith.science/pith/OTK53WRS5WWGEKM6G3PHYVUUWF/action/replication_record"}},"created_at":"2026-07-05T11:10:05.111711+00:00","updated_at":"2026-07-05T11:10:05.111711+00:00"}