{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XWYOYNBFLZVVL5UMDQV4W5NH7J","short_pith_number":"pith:XWYOYNBF","schema_version":"1.0","canonical_sha256":"bdb0ec34255e6b55f68c1c2bcb75a7fa5dc42aefecadc244912a4fc2eecad6ba","source":{"kind":"arxiv","id":"2411.02564","version":2},"attestation_state":"computed","paper":{"title":"Continual LLaVA: Continual Instruction Tuning in Large Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Henghui Ding, Ian Reid, Jiahua Dong, Meng Cao, Tiancai Wang, Xiangyu Zhang, Xiaodan Liang, Yingfei Liu, Yuyang Liu","submitted_at":"2024-11-04T19:55:32Z","abstract_excerpt":"Instruction tuning constitutes a prevalent technique for tailoring Large Vision Language Models (LVLMs) to meet individual task requirements. To date, most of the existing approaches are confined to single-task adaptation, whereas the requirements in real-world scenarios are inherently varied and continually evolving. Thus an ideal LVLM should sustain continual instruction tuning in the face of stream-task distributions (i.e., different domains, emerging capabilities, and new datasets) while minimizing the forgetting of previously acquired knowledge. To achieve this, we propose a new benchmark"},"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":"2411.02564","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-11-04T19:55:32Z","cross_cats_sorted":[],"title_canon_sha256":"feaddea3892d8ab7cc5890b437dca389967a3e0b19807810d9eb09098aee385c","abstract_canon_sha256":"6c3ab3295d7bed20546091a132052aaf4242234a8d9efb0a3479ef44531a0571"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:33:40.309786Z","signature_b64":"agHW4NiqQaTtYcqEXMNltUddWLNZfyDzP0RWkn/GEUNDlIrx1eqX5Z7CqVGT68HlKUPADe91UZ7asDuSsNS3AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bdb0ec34255e6b55f68c1c2bcb75a7fa5dc42aefecadc244912a4fc2eecad6ba","last_reissued_at":"2026-07-05T09:33:40.309152Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:33:40.309152Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Continual LLaVA: Continual Instruction Tuning in Large Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Henghui Ding, Ian Reid, Jiahua Dong, Meng Cao, Tiancai Wang, Xiangyu Zhang, Xiaodan Liang, Yingfei Liu, Yuyang Liu","submitted_at":"2024-11-04T19:55:32Z","abstract_excerpt":"Instruction tuning constitutes a prevalent technique for tailoring Large Vision Language Models (LVLMs) to meet individual task requirements. To date, most of the existing approaches are confined to single-task adaptation, whereas the requirements in real-world scenarios are inherently varied and continually evolving. Thus an ideal LVLM should sustain continual instruction tuning in the face of stream-task distributions (i.e., different domains, emerging capabilities, and new datasets) while minimizing the forgetting of previously acquired knowledge. To achieve this, we propose a new benchmark"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.02564","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/2411.02564/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":"2411.02564","created_at":"2026-07-05T09:33:40.309214+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.02564v2","created_at":"2026-07-05T09:33:40.309214+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.02564","created_at":"2026-07-05T09:33:40.309214+00:00"},{"alias_kind":"pith_short_12","alias_value":"XWYOYNBFLZVV","created_at":"2026-07-05T09:33:40.309214+00:00"},{"alias_kind":"pith_short_16","alias_value":"XWYOYNBFLZVVL5UM","created_at":"2026-07-05T09:33:40.309214+00:00"},{"alias_kind":"pith_short_8","alias_value":"XWYOYNBF","created_at":"2026-07-05T09:33:40.309214+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.24963","citing_title":"Curvature-Guided Mixing for MLLM Adaptation","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02020","citing_title":"Hidden Forgetting in Continual Multimodal Learning: When Accuracy Survives but Grounding Fails","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.12633","citing_title":"ECA: Efficient Continual Alignment for Open-Ended Image-to-Text Generation","ref_index":92,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10887","citing_title":"Listen, Look, and Learn: Learning Without Forgetting through SAM-Audio","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02502","citing_title":"CRAM: Centroid-Routing and Adaptive MoE for Multimodal Continual Instruction Tuning","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02576","citing_title":"ProtoAda: Prototype-Guided Adaptive Adapter Expansion and Geometric Consolidation for Multimodal Continual Instruction Tuning","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14938","citing_title":"Octopus: History-Free Gradient Orthogonalization for Continual Learning in Multimodal Large Language Models","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10765","citing_title":"Dynamic Cross-Modal Prompt Generation for Multimodal Continual Instruction Tuning","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04857","citing_title":"The Blind Spot of Adaptation: Quantifying and Mitigating Forgetting in Fine-tuned Driving Models","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XWYOYNBFLZVVL5UMDQV4W5NH7J","json":"https://pith.science/pith/XWYOYNBFLZVVL5UMDQV4W5NH7J.json","graph_json":"https://pith.science/api/pith-number/XWYOYNBFLZVVL5UMDQV4W5NH7J/graph.json","events_json":"https://pith.science/api/pith-number/XWYOYNBFLZVVL5UMDQV4W5NH7J/events.json","paper":"https://pith.science/paper/XWYOYNBF"},"agent_actions":{"view_html":"https://pith.science/pith/XWYOYNBFLZVVL5UMDQV4W5NH7J","download_json":"https://pith.science/pith/XWYOYNBFLZVVL5UMDQV4W5NH7J.json","view_paper":"https://pith.science/paper/XWYOYNBF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.02564&json=true","fetch_graph":"https://pith.science/api/pith-number/XWYOYNBFLZVVL5UMDQV4W5NH7J/graph.json","fetch_events":"https://pith.science/api/pith-number/XWYOYNBFLZVVL5UMDQV4W5NH7J/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XWYOYNBFLZVVL5UMDQV4W5NH7J/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XWYOYNBFLZVVL5UMDQV4W5NH7J/action/storage_attestation","attest_author":"https://pith.science/pith/XWYOYNBFLZVVL5UMDQV4W5NH7J/action/author_attestation","sign_citation":"https://pith.science/pith/XWYOYNBFLZVVL5UMDQV4W5NH7J/action/citation_signature","submit_replication":"https://pith.science/pith/XWYOYNBFLZVVL5UMDQV4W5NH7J/action/replication_record"}},"created_at":"2026-07-05T09:33:40.309214+00:00","updated_at":"2026-07-05T09:33:40.309214+00:00"}