{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:V3O2HPQTLEBRC6WX4PEHPUTMQV","short_pith_number":"pith:V3O2HPQT","schema_version":"1.0","canonical_sha256":"aedda3be135903117ad7e3c877d26c8570383ac3b10944760a5d78d497305ed3","source":{"kind":"arxiv","id":"2410.18035","version":1},"attestation_state":"computed","paper":{"title":"MiLoRA: Efficient Mixture of Low-Rank Adaptation for Large Language Models Fine-tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dan Chen, Huanran Zheng, Jingfan Zhang, Wei Zhu, Xing Tian, Yi Zhao","submitted_at":"2024-10-23T17:04:40Z","abstract_excerpt":"Low-rank adaptation (LoRA) and its mixture-of-experts (MOE) variants are highly effective parameter-efficient fine-tuning (PEFT) methods. However, they introduce significant latency in multi-tenant settings due to the LoRA modules and MOE routers added to multiple linear modules in the Transformer layer. To address this issue, we propose Mixture of Low-Rank Adaptation (MiLoRA), a novel and efficient LoRA variant. MiLoRA differs from previous MOE-style LoRA methods by considering each LoRA module as an expert and employing a prompt-aware routing mechanism. This mechanism calculates expert routi"},"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":"2410.18035","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-23T17:04:40Z","cross_cats_sorted":[],"title_canon_sha256":"8be6f36f9695a995b96192a997b75fe1920a093830d1fd084913dec494a40c7f","abstract_canon_sha256":"bb5b90c97a081c0402cbabc82e2e012c4cd038a2e18b5a9c954dcb68e52f3349"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:24:48.427471Z","signature_b64":"0n9m7kDdCBBxig9KS9f0U2bFwNo7gcq/gP8c5b47xawF77Wl/ZBwrUaI/gJzNEKuw4IXspGyt2O63QBGKEyJBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aedda3be135903117ad7e3c877d26c8570383ac3b10944760a5d78d497305ed3","last_reissued_at":"2026-07-05T09:24:48.426974Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:24:48.426974Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MiLoRA: Efficient Mixture of Low-Rank Adaptation for Large Language Models Fine-tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dan Chen, Huanran Zheng, Jingfan Zhang, Wei Zhu, Xing Tian, Yi Zhao","submitted_at":"2024-10-23T17:04:40Z","abstract_excerpt":"Low-rank adaptation (LoRA) and its mixture-of-experts (MOE) variants are highly effective parameter-efficient fine-tuning (PEFT) methods. However, they introduce significant latency in multi-tenant settings due to the LoRA modules and MOE routers added to multiple linear modules in the Transformer layer. To address this issue, we propose Mixture of Low-Rank Adaptation (MiLoRA), a novel and efficient LoRA variant. MiLoRA differs from previous MOE-style LoRA methods by considering each LoRA module as an expert and employing a prompt-aware routing mechanism. This mechanism calculates expert routi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.18035","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/2410.18035/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":"2410.18035","created_at":"2026-07-05T09:24:48.427031+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.18035v1","created_at":"2026-07-05T09:24:48.427031+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.18035","created_at":"2026-07-05T09:24:48.427031+00:00"},{"alias_kind":"pith_short_12","alias_value":"V3O2HPQTLEBR","created_at":"2026-07-05T09:24:48.427031+00:00"},{"alias_kind":"pith_short_16","alias_value":"V3O2HPQTLEBRC6WX","created_at":"2026-07-05T09:24:48.427031+00:00"},{"alias_kind":"pith_short_8","alias_value":"V3O2HPQT","created_at":"2026-07-05T09:24:48.427031+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.21035","citing_title":"Little by Little: Continual Learning via Incremental Mixture of Rank-1 Associative Memory Experts","ref_index":87,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V3O2HPQTLEBRC6WX4PEHPUTMQV","json":"https://pith.science/pith/V3O2HPQTLEBRC6WX4PEHPUTMQV.json","graph_json":"https://pith.science/api/pith-number/V3O2HPQTLEBRC6WX4PEHPUTMQV/graph.json","events_json":"https://pith.science/api/pith-number/V3O2HPQTLEBRC6WX4PEHPUTMQV/events.json","paper":"https://pith.science/paper/V3O2HPQT"},"agent_actions":{"view_html":"https://pith.science/pith/V3O2HPQTLEBRC6WX4PEHPUTMQV","download_json":"https://pith.science/pith/V3O2HPQTLEBRC6WX4PEHPUTMQV.json","view_paper":"https://pith.science/paper/V3O2HPQT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.18035&json=true","fetch_graph":"https://pith.science/api/pith-number/V3O2HPQTLEBRC6WX4PEHPUTMQV/graph.json","fetch_events":"https://pith.science/api/pith-number/V3O2HPQTLEBRC6WX4PEHPUTMQV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V3O2HPQTLEBRC6WX4PEHPUTMQV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V3O2HPQTLEBRC6WX4PEHPUTMQV/action/storage_attestation","attest_author":"https://pith.science/pith/V3O2HPQTLEBRC6WX4PEHPUTMQV/action/author_attestation","sign_citation":"https://pith.science/pith/V3O2HPQTLEBRC6WX4PEHPUTMQV/action/citation_signature","submit_replication":"https://pith.science/pith/V3O2HPQTLEBRC6WX4PEHPUTMQV/action/replication_record"}},"created_at":"2026-07-05T09:24:48.427031+00:00","updated_at":"2026-07-05T09:24:48.427031+00:00"}