{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:SEKRBCZZIJ4JO7N6HFLX4F4MWY","short_pith_number":"pith:SEKRBCZZ","schema_version":"1.0","canonical_sha256":"9115108b394278977dbe39577e178cb60acb399f23d13043100bda512687b624","source":{"kind":"arxiv","id":"2311.02684","version":3},"attestation_state":"computed","paper":{"title":"Octavius: Mitigating Task Interference in MLLMs via LoRA-MoE","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Huayang Liu, Jing Shao, Lu Sheng, Si Liu, Wanli Ouyang, Yu Qiao, Zeren Chen, Zhenfei Yin, Zhen Wang, Ziqin Wang","submitted_at":"2023-11-05T15:48:29Z","abstract_excerpt":"Recent studies have demonstrated Large Language Models (LLMs) can extend their zero-shot generalization capabilities to multimodal learning through instruction tuning. As more modalities and downstream tasks are introduced, negative conflicts and interference may have a worse impact on performance. While this phenomenon has been overlooked in previous work, we propose a novel and extensible framework, called Octavius, for comprehensive studies and experimentation on multimodal learning with Multimodal Large Language Models (MLLMs). Specifically, we combine the well-known Mixture-of-Experts (Mo"},"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":"2311.02684","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-11-05T15:48:29Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"ed0cfcd4ba4bd4979268f21e6c88b2769e6cfc8a1970addc9d39c77f7d4e89c4","abstract_canon_sha256":"c43589258cd29e38051ae633fc694cca2445c14b7aa5eed022eb93d66e1f484f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:39:51.684413Z","signature_b64":"iiPO6W8YeUTWJha7XInSGNv2HIVYCh6QWb/wKdbYgG/oZx2v5dYNjQ87l/xU8Da5awEr/0Jh/6c0gYCgxQGUDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9115108b394278977dbe39577e178cb60acb399f23d13043100bda512687b624","last_reissued_at":"2026-07-05T09:39:51.683871Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:39:51.683871Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Octavius: Mitigating Task Interference in MLLMs via LoRA-MoE","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Huayang Liu, Jing Shao, Lu Sheng, Si Liu, Wanli Ouyang, Yu Qiao, Zeren Chen, Zhenfei Yin, Zhen Wang, Ziqin Wang","submitted_at":"2023-11-05T15:48:29Z","abstract_excerpt":"Recent studies have demonstrated Large Language Models (LLMs) can extend their zero-shot generalization capabilities to multimodal learning through instruction tuning. As more modalities and downstream tasks are introduced, negative conflicts and interference may have a worse impact on performance. While this phenomenon has been overlooked in previous work, we propose a novel and extensible framework, called Octavius, for comprehensive studies and experimentation on multimodal learning with Multimodal Large Language Models (MLLMs). Specifically, we combine the well-known Mixture-of-Experts (Mo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.02684","kind":"arxiv","version":3},"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/2311.02684/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":"2311.02684","created_at":"2026-07-05T09:39:51.683940+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.02684v3","created_at":"2026-07-05T09:39:51.683940+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.02684","created_at":"2026-07-05T09:39:51.683940+00:00"},{"alias_kind":"pith_short_12","alias_value":"SEKRBCZZIJ4J","created_at":"2026-07-05T09:39:51.683940+00:00"},{"alias_kind":"pith_short_16","alias_value":"SEKRBCZZIJ4JO7N6","created_at":"2026-07-05T09:39:51.683940+00:00"},{"alias_kind":"pith_short_8","alias_value":"SEKRBCZZ","created_at":"2026-07-05T09:39:51.683940+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26287","citing_title":"GeMoE: Gating Entropy is All You Need for Uncertainty-aware Adaptive Routing in MoE-based Large Vision-Language Models","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2506.21035","citing_title":"Little by Little: Continual Learning via Incremental Mixture of Rank-1 Associative Memory Experts","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2507.00029","citing_title":"LoRA-Mixer: Coordinate Modular LoRA Experts Through Serial Attention Routing","ref_index":46,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SEKRBCZZIJ4JO7N6HFLX4F4MWY","json":"https://pith.science/pith/SEKRBCZZIJ4JO7N6HFLX4F4MWY.json","graph_json":"https://pith.science/api/pith-number/SEKRBCZZIJ4JO7N6HFLX4F4MWY/graph.json","events_json":"https://pith.science/api/pith-number/SEKRBCZZIJ4JO7N6HFLX4F4MWY/events.json","paper":"https://pith.science/paper/SEKRBCZZ"},"agent_actions":{"view_html":"https://pith.science/pith/SEKRBCZZIJ4JO7N6HFLX4F4MWY","download_json":"https://pith.science/pith/SEKRBCZZIJ4JO7N6HFLX4F4MWY.json","view_paper":"https://pith.science/paper/SEKRBCZZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.02684&json=true","fetch_graph":"https://pith.science/api/pith-number/SEKRBCZZIJ4JO7N6HFLX4F4MWY/graph.json","fetch_events":"https://pith.science/api/pith-number/SEKRBCZZIJ4JO7N6HFLX4F4MWY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SEKRBCZZIJ4JO7N6HFLX4F4MWY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SEKRBCZZIJ4JO7N6HFLX4F4MWY/action/storage_attestation","attest_author":"https://pith.science/pith/SEKRBCZZIJ4JO7N6HFLX4F4MWY/action/author_attestation","sign_citation":"https://pith.science/pith/SEKRBCZZIJ4JO7N6HFLX4F4MWY/action/citation_signature","submit_replication":"https://pith.science/pith/SEKRBCZZIJ4JO7N6HFLX4F4MWY/action/replication_record"}},"created_at":"2026-07-05T09:39:51.683940+00:00","updated_at":"2026-07-05T09:39:51.683940+00:00"}