{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:VPJMYUPODEVUDW6DSBH7C7FDMH","short_pith_number":"pith:VPJMYUPO","schema_version":"1.0","canonical_sha256":"abd2cc51ee192b41dbc3904ff17ca361c3aacbf21eb9f18d8b9f102bca545bac","source":{"kind":"arxiv","id":"2310.19182","version":1},"attestation_state":"computed","paper":{"title":"Fast Trainable Projection for Robust Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"James Seale Smith, Junjiao Tian, Yen-Cheng Liu, Zsolt Kira","submitted_at":"2023-10-29T22:52:43Z","abstract_excerpt":"Robust fine-tuning aims to achieve competitive in-distribution (ID) performance while maintaining the out-of-distribution (OOD) robustness of a pre-trained model when transferring it to a downstream task. Recently, projected gradient descent has been successfully used in robust fine-tuning by constraining the deviation from the initialization of the fine-tuned model explicitly through projection. However, algorithmically, two limitations prevent this method from being adopted more widely, scalability and efficiency. In this paper, we propose a new projection-based fine-tuning algorithm, Fast T"},"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":"2310.19182","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-10-29T22:52:43Z","cross_cats_sorted":[],"title_canon_sha256":"e9698510e59305e0cfc8b567287d298cc6feff35e558d3fdb3a097c1d396c74f","abstract_canon_sha256":"40702e3164cb53c3452f932fc962f7bf283f77f96636f96e4b0f74af7a138fbc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:06:47.246626Z","signature_b64":"1DpmeHUhySeaU6wh4FQnz0G/bGPeFM0uLiljiCHjgvuES5XZwD47MA3TYUbb/RJRi9vU1HVpWh2dqjUv0T+LDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"abd2cc51ee192b41dbc3904ff17ca361c3aacbf21eb9f18d8b9f102bca545bac","last_reissued_at":"2026-07-05T07:06:47.246166Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:06:47.246166Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fast Trainable Projection for Robust Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"James Seale Smith, Junjiao Tian, Yen-Cheng Liu, Zsolt Kira","submitted_at":"2023-10-29T22:52:43Z","abstract_excerpt":"Robust fine-tuning aims to achieve competitive in-distribution (ID) performance while maintaining the out-of-distribution (OOD) robustness of a pre-trained model when transferring it to a downstream task. Recently, projected gradient descent has been successfully used in robust fine-tuning by constraining the deviation from the initialization of the fine-tuned model explicitly through projection. However, algorithmically, two limitations prevent this method from being adopted more widely, scalability and efficiency. In this paper, we propose a new projection-based fine-tuning algorithm, Fast T"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.19182","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/2310.19182/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":"2310.19182","created_at":"2026-07-05T07:06:47.246221+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.19182v1","created_at":"2026-07-05T07:06:47.246221+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.19182","created_at":"2026-07-05T07:06:47.246221+00:00"},{"alias_kind":"pith_short_12","alias_value":"VPJMYUPODEVU","created_at":"2026-07-05T07:06:47.246221+00:00"},{"alias_kind":"pith_short_16","alias_value":"VPJMYUPODEVUDW6D","created_at":"2026-07-05T07:06:47.246221+00:00"},{"alias_kind":"pith_short_8","alias_value":"VPJMYUPO","created_at":"2026-07-05T07:06:47.246221+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/VPJMYUPODEVUDW6DSBH7C7FDMH","json":"https://pith.science/pith/VPJMYUPODEVUDW6DSBH7C7FDMH.json","graph_json":"https://pith.science/api/pith-number/VPJMYUPODEVUDW6DSBH7C7FDMH/graph.json","events_json":"https://pith.science/api/pith-number/VPJMYUPODEVUDW6DSBH7C7FDMH/events.json","paper":"https://pith.science/paper/VPJMYUPO"},"agent_actions":{"view_html":"https://pith.science/pith/VPJMYUPODEVUDW6DSBH7C7FDMH","download_json":"https://pith.science/pith/VPJMYUPODEVUDW6DSBH7C7FDMH.json","view_paper":"https://pith.science/paper/VPJMYUPO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.19182&json=true","fetch_graph":"https://pith.science/api/pith-number/VPJMYUPODEVUDW6DSBH7C7FDMH/graph.json","fetch_events":"https://pith.science/api/pith-number/VPJMYUPODEVUDW6DSBH7C7FDMH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VPJMYUPODEVUDW6DSBH7C7FDMH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VPJMYUPODEVUDW6DSBH7C7FDMH/action/storage_attestation","attest_author":"https://pith.science/pith/VPJMYUPODEVUDW6DSBH7C7FDMH/action/author_attestation","sign_citation":"https://pith.science/pith/VPJMYUPODEVUDW6DSBH7C7FDMH/action/citation_signature","submit_replication":"https://pith.science/pith/VPJMYUPODEVUDW6DSBH7C7FDMH/action/replication_record"}},"created_at":"2026-07-05T07:06:47.246221+00:00","updated_at":"2026-07-05T07:06:47.246221+00:00"}