{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:CZNJS46TGCZAFCDEO6GAFRRHRS","short_pith_number":"pith:CZNJS46T","schema_version":"1.0","canonical_sha256":"165a9973d330b2028864778c02c6278c98b00fdabfa2987107a1034a372d91b5","source":{"kind":"arxiv","id":"2508.04227","version":2},"attestation_state":"computed","paper":{"title":"Continual Learning for VLMs: A Survey and Taxonomy Beyond Forgetting","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Alexandra Gomez-Villa, Dipam Goswami, Joost Van De Weijer, Linlan Huang, Qiuhe Hong, Xialei Liu, Yonghong Tian, Yuyang Liu","submitted_at":"2025-08-06T09:03:10Z","abstract_excerpt":"Vision-language models (VLMs) and the recent surge of Multimodal Large Language Models (MLLMs) have revolutionized artificial intelligence with unprecedented cross-modal alignment and zero-shot generalization. However, enabling them to learn continually from non-stationary data remains a major challenge, as their cross-modal alignment and generalization capabilities are particularly vulnerable to catastrophic forgetting. Unlike traditional unimodal continual learning (CL), VLMs face unique challenges such as cross-modal feature drift, parameter interference due to shared architectures, and zer"},"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":"2508.04227","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2025-08-06T09:03:10Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"dec47ffcee357ab6f11f26ce337ad22c1a12d46c2cf4bf17d2494f289baf2455","abstract_canon_sha256":"3af2cb538ba47d7a8846cc2ee195f83d4f58cddafd51b792588760fa363ce34b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:52.751120Z","signature_b64":"MtSzgCrVbvkcamlV2Sm/FGLta85HqcvCO/tvDZ+bFnBVNTwryhUwJ664XZuoNRu+C8Qz8qcjw2xgOgx1Jg5zDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"165a9973d330b2028864778c02c6278c98b00fdabfa2987107a1034a372d91b5","last_reissued_at":"2026-05-20T00:02:52.750142Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:52.750142Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Continual Learning for VLMs: A Survey and Taxonomy Beyond Forgetting","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Alexandra Gomez-Villa, Dipam Goswami, Joost Van De Weijer, Linlan Huang, Qiuhe Hong, Xialei Liu, Yonghong Tian, Yuyang Liu","submitted_at":"2025-08-06T09:03:10Z","abstract_excerpt":"Vision-language models (VLMs) and the recent surge of Multimodal Large Language Models (MLLMs) have revolutionized artificial intelligence with unprecedented cross-modal alignment and zero-shot generalization. However, enabling them to learn continually from non-stationary data remains a major challenge, as their cross-modal alignment and generalization capabilities are particularly vulnerable to catastrophic forgetting. Unlike traditional unimodal continual learning (CL), VLMs face unique challenges such as cross-modal feature drift, parameter interference due to shared architectures, and zer"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.04227","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/2508.04227/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":"2508.04227","created_at":"2026-05-20T00:02:52.750270+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.04227v2","created_at":"2026-05-20T00:02:52.750270+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.04227","created_at":"2026-05-20T00:02:52.750270+00:00"},{"alias_kind":"pith_short_12","alias_value":"CZNJS46TGCZA","created_at":"2026-05-20T00:02:52.750270+00:00"},{"alias_kind":"pith_short_16","alias_value":"CZNJS46TGCZAFCDE","created_at":"2026-05-20T00:02:52.750270+00:00"},{"alias_kind":"pith_short_8","alias_value":"CZNJS46T","created_at":"2026-05-20T00:02:52.750270+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":6,"sample":[{"citing_arxiv_id":"2605.18903","citing_title":"Reasoning Portability: Guiding Continual Learning for MLLMs in the RLVR Era","ref_index":28,"is_internal_anchor":true},{"citing_arxiv_id":"2605.19301","citing_title":"iGSP:Implicit Gradient Subspace Projection for Efficient Continual Learning of Vision-Language Models","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2604.07965","citing_title":"DSCA: Dynamic Subspace Concept Alignment for Lifelong VLM Editing","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2604.14016","citing_title":"MAny: Merge Anything for Multimodal Continual Instruction Tuning","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2604.14779","citing_title":"AIM: Asymmetric Information Masking for Visual Question Answering Continual Learning","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2604.21280","citing_title":"ImageHD: Energy-Efficient On-Device Continual Learning of Visual Representations via Hyperdimensional Computing","ref_index":29,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CZNJS46TGCZAFCDEO6GAFRRHRS","json":"https://pith.science/pith/CZNJS46TGCZAFCDEO6GAFRRHRS.json","graph_json":"https://pith.science/api/pith-number/CZNJS46TGCZAFCDEO6GAFRRHRS/graph.json","events_json":"https://pith.science/api/pith-number/CZNJS46TGCZAFCDEO6GAFRRHRS/events.json","paper":"https://pith.science/paper/CZNJS46T"},"agent_actions":{"view_html":"https://pith.science/pith/CZNJS46TGCZAFCDEO6GAFRRHRS","download_json":"https://pith.science/pith/CZNJS46TGCZAFCDEO6GAFRRHRS.json","view_paper":"https://pith.science/paper/CZNJS46T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.04227&json=true","fetch_graph":"https://pith.science/api/pith-number/CZNJS46TGCZAFCDEO6GAFRRHRS/graph.json","fetch_events":"https://pith.science/api/pith-number/CZNJS46TGCZAFCDEO6GAFRRHRS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CZNJS46TGCZAFCDEO6GAFRRHRS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CZNJS46TGCZAFCDEO6GAFRRHRS/action/storage_attestation","attest_author":"https://pith.science/pith/CZNJS46TGCZAFCDEO6GAFRRHRS/action/author_attestation","sign_citation":"https://pith.science/pith/CZNJS46TGCZAFCDEO6GAFRRHRS/action/citation_signature","submit_replication":"https://pith.science/pith/CZNJS46TGCZAFCDEO6GAFRRHRS/action/replication_record"}},"created_at":"2026-05-20T00:02:52.750270+00:00","updated_at":"2026-05-20T00:02:52.750270+00:00"}