{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:75URF6A7LWZGMKXJPR3Z6BJKFH","short_pith_number":"pith:75URF6A7","schema_version":"1.0","canonical_sha256":"ff6912f81f5db2662ae97c779f052a29fb1078888332b6b8c58df28d92eed135","source":{"kind":"arxiv","id":"2411.02207","version":1},"attestation_state":"computed","paper":{"title":"Collective Model Intelligence Requires Compatible Specialization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jyothish Pari, Pulkit Agrawal, Samy Jelassi","submitted_at":"2024-11-04T15:59:16Z","abstract_excerpt":"In this work, we explore the limitations of combining models by averaging intermediate features, referred to as model merging, and propose a new direction for achieving collective model intelligence through what we call compatible specialization. Current methods for model merging, such as parameter and feature averaging, struggle to effectively combine specialized models due to representational divergence during fine-tuning. As models specialize to their individual domains, their internal feature representations become increasingly incompatible, leading to poor performance when attempting to m"},"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":"2411.02207","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T15:59:16Z","cross_cats_sorted":[],"title_canon_sha256":"51b1f4b399f4ffb7fcd87e831e352ce5a832da184e823adc744b2850563a8ca5","abstract_canon_sha256":"cdd525fdbe72317e39c4837a38dfd495b02bc18b5d3672fde48123cafee4f48c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:30:43.390188Z","signature_b64":"HjNmOT8say9z0J8hIBfNhjk9WZd1mjaDCDXbbTu8RCp2EGlbtGd0mpZuj8s1R2yGRWJlBt81qlYtariFJgMoDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ff6912f81f5db2662ae97c779f052a29fb1078888332b6b8c58df28d92eed135","last_reissued_at":"2026-07-05T09:30:43.389743Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:30:43.389743Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Collective Model Intelligence Requires Compatible Specialization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jyothish Pari, Pulkit Agrawal, Samy Jelassi","submitted_at":"2024-11-04T15:59:16Z","abstract_excerpt":"In this work, we explore the limitations of combining models by averaging intermediate features, referred to as model merging, and propose a new direction for achieving collective model intelligence through what we call compatible specialization. Current methods for model merging, such as parameter and feature averaging, struggle to effectively combine specialized models due to representational divergence during fine-tuning. As models specialize to their individual domains, their internal feature representations become increasingly incompatible, leading to poor performance when attempting to m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.02207","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/2411.02207/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":"2411.02207","created_at":"2026-07-05T09:30:43.389799+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.02207v1","created_at":"2026-07-05T09:30:43.389799+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.02207","created_at":"2026-07-05T09:30:43.389799+00:00"},{"alias_kind":"pith_short_12","alias_value":"75URF6A7LWZG","created_at":"2026-07-05T09:30:43.389799+00:00"},{"alias_kind":"pith_short_16","alias_value":"75URF6A7LWZGMKXJ","created_at":"2026-07-05T09:30:43.389799+00:00"},{"alias_kind":"pith_short_8","alias_value":"75URF6A7","created_at":"2026-07-05T09:30:43.389799+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.07024","citing_title":"FlexOlmo: Open Language Models for Flexible Data Use","ref_index":43,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/75URF6A7LWZGMKXJPR3Z6BJKFH","json":"https://pith.science/pith/75URF6A7LWZGMKXJPR3Z6BJKFH.json","graph_json":"https://pith.science/api/pith-number/75URF6A7LWZGMKXJPR3Z6BJKFH/graph.json","events_json":"https://pith.science/api/pith-number/75URF6A7LWZGMKXJPR3Z6BJKFH/events.json","paper":"https://pith.science/paper/75URF6A7"},"agent_actions":{"view_html":"https://pith.science/pith/75URF6A7LWZGMKXJPR3Z6BJKFH","download_json":"https://pith.science/pith/75URF6A7LWZGMKXJPR3Z6BJKFH.json","view_paper":"https://pith.science/paper/75URF6A7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.02207&json=true","fetch_graph":"https://pith.science/api/pith-number/75URF6A7LWZGMKXJPR3Z6BJKFH/graph.json","fetch_events":"https://pith.science/api/pith-number/75URF6A7LWZGMKXJPR3Z6BJKFH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/75URF6A7LWZGMKXJPR3Z6BJKFH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/75URF6A7LWZGMKXJPR3Z6BJKFH/action/storage_attestation","attest_author":"https://pith.science/pith/75URF6A7LWZGMKXJPR3Z6BJKFH/action/author_attestation","sign_citation":"https://pith.science/pith/75URF6A7LWZGMKXJPR3Z6BJKFH/action/citation_signature","submit_replication":"https://pith.science/pith/75URF6A7LWZGMKXJPR3Z6BJKFH/action/replication_record"}},"created_at":"2026-07-05T09:30:43.389799+00:00","updated_at":"2026-07-05T09:30:43.389799+00:00"}