{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BEV2FLEZVLXSRDBPH5M4BHTPOK","short_pith_number":"pith:BEV2FLEZ","schema_version":"1.0","canonical_sha256":"092ba2ac99aaef288c2f3f59c09e6f72a696e4cb3a2c0ac6e4b6e810d54b619c","source":{"kind":"arxiv","id":"2502.04098","version":2},"attestation_state":"computed","paper":{"title":"Efficient Few-Shot Continual Learning in Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Aristeidis Panos, Daniel Olmeda Reino, Rahaf Aljundi, Richard E. Turner","submitted_at":"2025-02-06T14:20:55Z","abstract_excerpt":"Vision-language models (VLMs) excel in tasks such as visual question answering and image captioning. However, VLMs are often limited by their use of pretrained image encoders, like CLIP, leading to image understanding errors that hinder overall performance. On top of that, real-world applications often require the model to be continuously adapted as new and often limited data continuously arrive. To address this, we propose LoRSU (Low-Rank Adaptation with Structured Updates), a robust and computationally efficient method for selectively updating image encoders within VLMs. LoRSU introduces str"},"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":"2502.04098","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-02-06T14:20:55Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"24344415300ae74daef185e35d4377674a0a51fb8f479331500c4f02fdea2f1a","abstract_canon_sha256":"2c8488e479d580a998a421725e53e46149e59244d70ec97a2d7af58cf68589e6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:10:49.595682Z","signature_b64":"KizrlT7hnbpQDZYqeQC5jhJ16S8fiYDtwi3YC13vYJR850egK/nA/1yVwTYh0f41BxpAM4ZAB8cKZe42v7fnDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"092ba2ac99aaef288c2f3f59c09e6f72a696e4cb3a2c0ac6e4b6e810d54b619c","last_reissued_at":"2026-07-05T10:10:49.595259Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:10:49.595259Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Efficient Few-Shot Continual Learning in Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Aristeidis Panos, Daniel Olmeda Reino, Rahaf Aljundi, Richard E. Turner","submitted_at":"2025-02-06T14:20:55Z","abstract_excerpt":"Vision-language models (VLMs) excel in tasks such as visual question answering and image captioning. However, VLMs are often limited by their use of pretrained image encoders, like CLIP, leading to image understanding errors that hinder overall performance. On top of that, real-world applications often require the model to be continuously adapted as new and often limited data continuously arrive. To address this, we propose LoRSU (Low-Rank Adaptation with Structured Updates), a robust and computationally efficient method for selectively updating image encoders within VLMs. LoRSU introduces str"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04098","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/2502.04098/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":"2502.04098","created_at":"2026-07-05T10:10:49.595328+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.04098v2","created_at":"2026-07-05T10:10:49.595328+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04098","created_at":"2026-07-05T10:10:49.595328+00:00"},{"alias_kind":"pith_short_12","alias_value":"BEV2FLEZVLXS","created_at":"2026-07-05T10:10:49.595328+00:00"},{"alias_kind":"pith_short_16","alias_value":"BEV2FLEZVLXSRDBP","created_at":"2026-07-05T10:10:49.595328+00:00"},{"alias_kind":"pith_short_8","alias_value":"BEV2FLEZ","created_at":"2026-07-05T10:10:49.595328+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/BEV2FLEZVLXSRDBPH5M4BHTPOK","json":"https://pith.science/pith/BEV2FLEZVLXSRDBPH5M4BHTPOK.json","graph_json":"https://pith.science/api/pith-number/BEV2FLEZVLXSRDBPH5M4BHTPOK/graph.json","events_json":"https://pith.science/api/pith-number/BEV2FLEZVLXSRDBPH5M4BHTPOK/events.json","paper":"https://pith.science/paper/BEV2FLEZ"},"agent_actions":{"view_html":"https://pith.science/pith/BEV2FLEZVLXSRDBPH5M4BHTPOK","download_json":"https://pith.science/pith/BEV2FLEZVLXSRDBPH5M4BHTPOK.json","view_paper":"https://pith.science/paper/BEV2FLEZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.04098&json=true","fetch_graph":"https://pith.science/api/pith-number/BEV2FLEZVLXSRDBPH5M4BHTPOK/graph.json","fetch_events":"https://pith.science/api/pith-number/BEV2FLEZVLXSRDBPH5M4BHTPOK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BEV2FLEZVLXSRDBPH5M4BHTPOK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BEV2FLEZVLXSRDBPH5M4BHTPOK/action/storage_attestation","attest_author":"https://pith.science/pith/BEV2FLEZVLXSRDBPH5M4BHTPOK/action/author_attestation","sign_citation":"https://pith.science/pith/BEV2FLEZVLXSRDBPH5M4BHTPOK/action/citation_signature","submit_replication":"https://pith.science/pith/BEV2FLEZVLXSRDBPH5M4BHTPOK/action/replication_record"}},"created_at":"2026-07-05T10:10:49.595328+00:00","updated_at":"2026-07-05T10:10:49.595328+00:00"}