{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:EOB54FRIR6EOBGRGKJWIXITUSP","short_pith_number":"pith:EOB54FRI","schema_version":"1.0","canonical_sha256":"2383de16288f88e09a26526c8ba27493e42c9c6a89f493ba324c2365562b299c","source":{"kind":"arxiv","id":"2505.21226","version":2},"attestation_state":"computed","paper":{"title":"Why Do More Experts Fail? A Theoretical Analysis of Model Merging","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Daling Wang, Hinrich Sch\\\"utze, Peiqin Lin, Shi Feng, Xiaocui Yang, Xingle Xu, Yiqun Zhang, Yongkang Liu, Zijing Wang","submitted_at":"2025-05-27T14:10:46Z","abstract_excerpt":"Model merging dramatically reduces storage and computational resources by combining multiple expert models into a single multi-task model. Although recent model merging methods have shown promising results, they struggle to maintain performance gains as the number of merged models increases. In this paper, we investigate the key obstacles that limit the scalability of model merging when integrating a large number of expert models. First, we prove that there is an upper bound on model merging. Further theoretical analysis reveals that the limited effective parameter space imposes a strict const"},"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":"2505.21226","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-27T14:10:46Z","cross_cats_sorted":[],"title_canon_sha256":"7d6c30ce360af68936c910953311f4e59049e33db6eba8b36bb447909b8c496f","abstract_canon_sha256":"b3f3b524550443535b7161b44940e52cde942bd973572044845fe56f55fca6ee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:15:00.464816Z","signature_b64":"y9osH633mniUrAaGGm8LCdriquz/txjzYuA3Dm1BQez430orilG1XyM6notZ0PyQIildmxiaSEht0BYi8wy3AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2383de16288f88e09a26526c8ba27493e42c9c6a89f493ba324c2365562b299c","last_reissued_at":"2026-07-05T11:15:00.464347Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:15:00.464347Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Why Do More Experts Fail? A Theoretical Analysis of Model Merging","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Daling Wang, Hinrich Sch\\\"utze, Peiqin Lin, Shi Feng, Xiaocui Yang, Xingle Xu, Yiqun Zhang, Yongkang Liu, Zijing Wang","submitted_at":"2025-05-27T14:10:46Z","abstract_excerpt":"Model merging dramatically reduces storage and computational resources by combining multiple expert models into a single multi-task model. Although recent model merging methods have shown promising results, they struggle to maintain performance gains as the number of merged models increases. In this paper, we investigate the key obstacles that limit the scalability of model merging when integrating a large number of expert models. First, we prove that there is an upper bound on model merging. Further theoretical analysis reveals that the limited effective parameter space imposes a strict const"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.21226","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/2505.21226/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":"2505.21226","created_at":"2026-07-05T11:15:00.464404+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.21226v2","created_at":"2026-07-05T11:15:00.464404+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.21226","created_at":"2026-07-05T11:15:00.464404+00:00"},{"alias_kind":"pith_short_12","alias_value":"EOB54FRIR6EO","created_at":"2026-07-05T11:15:00.464404+00:00"},{"alias_kind":"pith_short_16","alias_value":"EOB54FRIR6EOBGRG","created_at":"2026-07-05T11:15:00.464404+00:00"},{"alias_kind":"pith_short_8","alias_value":"EOB54FRI","created_at":"2026-07-05T11:15:00.464404+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.28373","citing_title":"Model Merging to Evolution: Parameter Space Exploration for Expert Models","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12960","citing_title":"DiM\\textsuperscript{3}: Bridging Multilingual and Multimodal Models via Direction- and Magnitude-Aware Merging","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2408.07666","citing_title":"Model Merging in LLMs, MLLMs, and Beyond: Methods, Theories, Applications and Opportunities","ref_index":245,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12960","citing_title":"DiM\\textsuperscript{3}: Bridging Multilingual and Multimodal Models via Direction- and Magnitude-Aware Merging","ref_index":46,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EOB54FRIR6EOBGRGKJWIXITUSP","json":"https://pith.science/pith/EOB54FRIR6EOBGRGKJWIXITUSP.json","graph_json":"https://pith.science/api/pith-number/EOB54FRIR6EOBGRGKJWIXITUSP/graph.json","events_json":"https://pith.science/api/pith-number/EOB54FRIR6EOBGRGKJWIXITUSP/events.json","paper":"https://pith.science/paper/EOB54FRI"},"agent_actions":{"view_html":"https://pith.science/pith/EOB54FRIR6EOBGRGKJWIXITUSP","download_json":"https://pith.science/pith/EOB54FRIR6EOBGRGKJWIXITUSP.json","view_paper":"https://pith.science/paper/EOB54FRI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.21226&json=true","fetch_graph":"https://pith.science/api/pith-number/EOB54FRIR6EOBGRGKJWIXITUSP/graph.json","fetch_events":"https://pith.science/api/pith-number/EOB54FRIR6EOBGRGKJWIXITUSP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EOB54FRIR6EOBGRGKJWIXITUSP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EOB54FRIR6EOBGRGKJWIXITUSP/action/storage_attestation","attest_author":"https://pith.science/pith/EOB54FRIR6EOBGRGKJWIXITUSP/action/author_attestation","sign_citation":"https://pith.science/pith/EOB54FRIR6EOBGRGKJWIXITUSP/action/citation_signature","submit_replication":"https://pith.science/pith/EOB54FRIR6EOBGRGKJWIXITUSP/action/replication_record"}},"created_at":"2026-07-05T11:15:00.464404+00:00","updated_at":"2026-07-05T11:15:00.464404+00:00"}