{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:TR5NRPWKPT5LD22ECFYJSXOF4S","short_pith_number":"pith:TR5NRPWK","schema_version":"1.0","canonical_sha256":"9c7ad8beca7cfab1eb441170995dc5e49c17f8fdf4e4620db85c652a57c09cf5","source":{"kind":"arxiv","id":"2501.10062","version":2},"attestation_state":"computed","paper":{"title":"OMoE: Diversifying Mixture of Low-Rank Adaptation by Orthogonal Finetuning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Dongmin Li, Huimu Wang, Jinyuan Feng, Tianyi Hu, Xiaolin Ai, Zhiqiang Pu","submitted_at":"2025-01-17T09:27:08Z","abstract_excerpt":"Building mixture-of-experts (MoE) architecture for Low-rank adaptation (LoRA) is emerging as a potential direction in parameter-efficient fine-tuning (PEFT) for its modular design and remarkable performance. However, simply stacking the number of experts cannot guarantee significant improvement. In this work, we first conduct qualitative analysis to indicate that experts collapse to similar representations in vanilla MoE, limiting the capacity of modular design and computational efficiency. Ulteriorly, Our analysis reveals that the performance of previous MoE variants maybe limited by a lack o"},"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":"2501.10062","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-17T09:27:08Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"b16346e68f0f26d785b53891381a6570661d851c3486be861156e04fb414df3f","abstract_canon_sha256":"f7ed275c32254c2b6532397f918c00901516469582f487eff3daefd56759f9b3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:40:32.388400Z","signature_b64":"C9MJ9Rdc4eoB5JwuWsz0NguC99SCAyaz7oxkpy74epBdZI0fwebfONn5Jm/pBr/uS7GLP3SPY4mB0eGSZy/NBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9c7ad8beca7cfab1eb441170995dc5e49c17f8fdf4e4620db85c652a57c09cf5","last_reissued_at":"2026-07-05T11:40:32.387920Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:40:32.387920Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"OMoE: Diversifying Mixture of Low-Rank Adaptation by Orthogonal Finetuning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Dongmin Li, Huimu Wang, Jinyuan Feng, Tianyi Hu, Xiaolin Ai, Zhiqiang Pu","submitted_at":"2025-01-17T09:27:08Z","abstract_excerpt":"Building mixture-of-experts (MoE) architecture for Low-rank adaptation (LoRA) is emerging as a potential direction in parameter-efficient fine-tuning (PEFT) for its modular design and remarkable performance. However, simply stacking the number of experts cannot guarantee significant improvement. In this work, we first conduct qualitative analysis to indicate that experts collapse to similar representations in vanilla MoE, limiting the capacity of modular design and computational efficiency. Ulteriorly, Our analysis reveals that the performance of previous MoE variants maybe limited by a lack o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.10062","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/2501.10062/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":"2501.10062","created_at":"2026-07-05T11:40:32.387966+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.10062v2","created_at":"2026-07-05T11:40:32.387966+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.10062","created_at":"2026-07-05T11:40:32.387966+00:00"},{"alias_kind":"pith_short_12","alias_value":"TR5NRPWKPT5L","created_at":"2026-07-05T11:40:32.387966+00:00"},{"alias_kind":"pith_short_16","alias_value":"TR5NRPWKPT5LD22E","created_at":"2026-07-05T11:40:32.387966+00:00"},{"alias_kind":"pith_short_8","alias_value":"TR5NRPWK","created_at":"2026-07-05T11:40:32.387966+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22878","citing_title":"Priority-Aware Learning-Unlearning Correction for Dynamic Decentralized LoRA Fine-Tuning","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03631","citing_title":"AnchorMoE: Interpretable Time Series Classification via Anchor-Routed MoE","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20247","citing_title":"CP-MoE: Consistency-Preserving Mixture-of-Experts for Continual Learning","ref_index":5,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TR5NRPWKPT5LD22ECFYJSXOF4S","json":"https://pith.science/pith/TR5NRPWKPT5LD22ECFYJSXOF4S.json","graph_json":"https://pith.science/api/pith-number/TR5NRPWKPT5LD22ECFYJSXOF4S/graph.json","events_json":"https://pith.science/api/pith-number/TR5NRPWKPT5LD22ECFYJSXOF4S/events.json","paper":"https://pith.science/paper/TR5NRPWK"},"agent_actions":{"view_html":"https://pith.science/pith/TR5NRPWKPT5LD22ECFYJSXOF4S","download_json":"https://pith.science/pith/TR5NRPWKPT5LD22ECFYJSXOF4S.json","view_paper":"https://pith.science/paper/TR5NRPWK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.10062&json=true","fetch_graph":"https://pith.science/api/pith-number/TR5NRPWKPT5LD22ECFYJSXOF4S/graph.json","fetch_events":"https://pith.science/api/pith-number/TR5NRPWKPT5LD22ECFYJSXOF4S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TR5NRPWKPT5LD22ECFYJSXOF4S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TR5NRPWKPT5LD22ECFYJSXOF4S/action/storage_attestation","attest_author":"https://pith.science/pith/TR5NRPWKPT5LD22ECFYJSXOF4S/action/author_attestation","sign_citation":"https://pith.science/pith/TR5NRPWKPT5LD22ECFYJSXOF4S/action/citation_signature","submit_replication":"https://pith.science/pith/TR5NRPWKPT5LD22ECFYJSXOF4S/action/replication_record"}},"created_at":"2026-07-05T11:40:32.387966+00:00","updated_at":"2026-07-05T11:40:32.387966+00:00"}