{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DXNPJORHKAJ2GNIU4MSIPAFD63","short_pith_number":"pith:DXNPJORH","schema_version":"1.0","canonical_sha256":"1ddaf4ba275013a33514e3248780a3f6d4a59550e432cbf0e4e0a5a3f5e22a1e","source":{"kind":"arxiv","id":"2408.01505","version":1},"attestation_state":"computed","paper":{"title":"MoDE: Effective Multi-task Parameter Efficient Fine-Tuning with a Mixture of Dyadic Experts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Abhinav Rastogi, Harsh Lara, Lin Ning, Meiqi Guo","submitted_at":"2024-08-02T18:05:10Z","abstract_excerpt":"Parameter-efficient fine-tuning techniques like Low-Rank Adaptation (LoRA) have revolutionized the adaptation of large language models (LLMs) to diverse tasks. Recent efforts have explored mixtures of LoRA modules for multi-task settings. However, our analysis reveals redundancy in the down-projection matrices of these architectures. This observation motivates our proposed method, Mixture of Dyadic Experts (MoDE), which introduces a novel design for efficient multi-task adaptation. This is done by sharing the down-projection matrix across tasks and employing atomic rank-one adapters, coupled w"},"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":"2408.01505","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-08-02T18:05:10Z","cross_cats_sorted":[],"title_canon_sha256":"c7396600d6131c4beac936f2d1b2755613f7dae0f6b5dacfd7c656d5f27b4880","abstract_canon_sha256":"4c91656bed1ba1c3a49306eb679f48e0fd89d5aeeefca0e699a91051e1339666"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:51:58.795016Z","signature_b64":"eON8F6/AfLVKnsgxZdGzGNqwC/5NambNLH/V7wgkBObq+tCyxywOCBnyUKXHeXzOeHDYOiPmbLjNqrqc60bIAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1ddaf4ba275013a33514e3248780a3f6d4a59550e432cbf0e4e0a5a3f5e22a1e","last_reissued_at":"2026-07-05T08:51:58.794604Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:51:58.794604Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MoDE: Effective Multi-task Parameter Efficient Fine-Tuning with a Mixture of Dyadic Experts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Abhinav Rastogi, Harsh Lara, Lin Ning, Meiqi Guo","submitted_at":"2024-08-02T18:05:10Z","abstract_excerpt":"Parameter-efficient fine-tuning techniques like Low-Rank Adaptation (LoRA) have revolutionized the adaptation of large language models (LLMs) to diverse tasks. Recent efforts have explored mixtures of LoRA modules for multi-task settings. However, our analysis reveals redundancy in the down-projection matrices of these architectures. This observation motivates our proposed method, Mixture of Dyadic Experts (MoDE), which introduces a novel design for efficient multi-task adaptation. This is done by sharing the down-projection matrix across tasks and employing atomic rank-one adapters, coupled w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.01505","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/2408.01505/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":"2408.01505","created_at":"2026-07-05T08:51:58.794667+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.01505v1","created_at":"2026-07-05T08:51:58.794667+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.01505","created_at":"2026-07-05T08:51:58.794667+00:00"},{"alias_kind":"pith_short_12","alias_value":"DXNPJORHKAJ2","created_at":"2026-07-05T08:51:58.794667+00:00"},{"alias_kind":"pith_short_16","alias_value":"DXNPJORHKAJ2GNIU","created_at":"2026-07-05T08:51:58.794667+00:00"},{"alias_kind":"pith_short_8","alias_value":"DXNPJORH","created_at":"2026-07-05T08:51:58.794667+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/DXNPJORHKAJ2GNIU4MSIPAFD63","json":"https://pith.science/pith/DXNPJORHKAJ2GNIU4MSIPAFD63.json","graph_json":"https://pith.science/api/pith-number/DXNPJORHKAJ2GNIU4MSIPAFD63/graph.json","events_json":"https://pith.science/api/pith-number/DXNPJORHKAJ2GNIU4MSIPAFD63/events.json","paper":"https://pith.science/paper/DXNPJORH"},"agent_actions":{"view_html":"https://pith.science/pith/DXNPJORHKAJ2GNIU4MSIPAFD63","download_json":"https://pith.science/pith/DXNPJORHKAJ2GNIU4MSIPAFD63.json","view_paper":"https://pith.science/paper/DXNPJORH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.01505&json=true","fetch_graph":"https://pith.science/api/pith-number/DXNPJORHKAJ2GNIU4MSIPAFD63/graph.json","fetch_events":"https://pith.science/api/pith-number/DXNPJORHKAJ2GNIU4MSIPAFD63/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DXNPJORHKAJ2GNIU4MSIPAFD63/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DXNPJORHKAJ2GNIU4MSIPAFD63/action/storage_attestation","attest_author":"https://pith.science/pith/DXNPJORHKAJ2GNIU4MSIPAFD63/action/author_attestation","sign_citation":"https://pith.science/pith/DXNPJORHKAJ2GNIU4MSIPAFD63/action/citation_signature","submit_replication":"https://pith.science/pith/DXNPJORHKAJ2GNIU4MSIPAFD63/action/replication_record"}},"created_at":"2026-07-05T08:51:58.794667+00:00","updated_at":"2026-07-05T08:51:58.794667+00:00"}