{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:XUFIRCUTFRYOYI3XPJVQCS7YUX","short_pith_number":"pith:XUFIRCUT","canonical_record":{"source":{"id":"2303.10720","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-19T17:30:44Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"5f1a0afb8b779833165ca4493be62cf743c44c2fc31df58b40c9005e76e0e73b","abstract_canon_sha256":"91dbad81ec0527d32a91e8ba8a5cabe74f4a1d471d8ada37612ddd627c4e3759"},"schema_version":"1.0"},"canonical_sha256":"bd0a888a932c70ec23777a6b014bf8a5e64af6801d4ac8cb66d374b4221b6dde","source":{"kind":"arxiv","id":"2303.10720","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2303.10720","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"arxiv_version","alias_value":"2303.10720v2","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.10720","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"pith_short_12","alias_value":"XUFIRCUTFRYO","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"pith_short_16","alias_value":"XUFIRCUTFRYOYI3X","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"pith_short_8","alias_value":"XUFIRCUT","created_at":"2026-07-05T05:55:22Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:XUFIRCUTFRYOYI3XPJVQCS7YUX","target":"record","payload":{"canonical_record":{"source":{"id":"2303.10720","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-19T17:30:44Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"5f1a0afb8b779833165ca4493be62cf743c44c2fc31df58b40c9005e76e0e73b","abstract_canon_sha256":"91dbad81ec0527d32a91e8ba8a5cabe74f4a1d471d8ada37612ddd627c4e3759"},"schema_version":"1.0"},"canonical_sha256":"bd0a888a932c70ec23777a6b014bf8a5e64af6801d4ac8cb66d374b4221b6dde","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:55:22.153106Z","signature_b64":"dBgQqZaFRkN1rKuh//Fa4s6mPl815Eorlt4nhx/4A9rnMFiy6W/dcvV5CtOR/q5KepAtBPaLQVTGXXtR+5iGAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bd0a888a932c70ec23777a6b014bf8a5e64af6801d4ac8cb66d374b4221b6dde","last_reissued_at":"2026-07-05T05:55:22.152727Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:55:22.152727Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2303.10720","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:55:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XhceRNjfsDvYD/lp3eXTpfBi4Ooh23VicSzZUi7lHQGoZzMRbdPTK7Tf7LZgiQ4FO1cCuZ7cYRCxsCfTzhAxBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T14:39:58.592036Z"},"content_sha256":"9e3d2a7bf041ad09a511cce6de0c8796b7775a4301851762d01b39ecd4bd3992","schema_version":"1.0","event_id":"sha256:9e3d2a7bf041ad09a511cce6de0c8796b7775a4301851762d01b39ecd4bd3992"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:XUFIRCUTFRYOYI3XPJVQCS7YUX","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Trainable Projected Gradient Method for Robust Fine-tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Chih-Yao Ma, Junjiao Tian, Xiaoliang Dai, Yen-Cheng Liu, Zecheng He, Zsolt Kira","submitted_at":"2023-03-19T17:30:44Z","abstract_excerpt":"Recent studies on transfer learning have shown that selectively fine-tuning a subset of layers or customizing different learning rates for each layer can greatly improve robustness to out-of-distribution (OOD) data and retain generalization capability in the pre-trained models. However, most of these methods employ manually crafted heuristics or expensive hyper-parameter searches, which prevent them from scaling up to large datasets and neural networks. To solve this problem, we propose Trainable Projected Gradient Method (TPGM) to automatically learn the constraint imposed for each layer for "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.10720","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/2303.10720/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:55:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"g2LcSiLKM/ylOAvelsGGMwAdB1siRCegfP0YSi6L91jnsPfpRXqJVXJmmmVzzyjvGtQ/EQioE/lBchDDa8x/Dg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T14:39:58.592973Z"},"content_sha256":"b1b074ea3ff2f44c3e34b039d9cb1b71c842e9b451a9f66cce437e6b380f136b","schema_version":"1.0","event_id":"sha256:b1b074ea3ff2f44c3e34b039d9cb1b71c842e9b451a9f66cce437e6b380f136b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XUFIRCUTFRYOYI3XPJVQCS7YUX/bundle.json","state_url":"https://pith.science/pith/XUFIRCUTFRYOYI3XPJVQCS7YUX/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XUFIRCUTFRYOYI3XPJVQCS7YUX/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-08T14:39:58Z","links":{"resolver":"https://pith.science/pith/XUFIRCUTFRYOYI3XPJVQCS7YUX","bundle":"https://pith.science/pith/XUFIRCUTFRYOYI3XPJVQCS7YUX/bundle.json","state":"https://pith.science/pith/XUFIRCUTFRYOYI3XPJVQCS7YUX/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XUFIRCUTFRYOYI3XPJVQCS7YUX/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:XUFIRCUTFRYOYI3XPJVQCS7YUX","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"91dbad81ec0527d32a91e8ba8a5cabe74f4a1d471d8ada37612ddd627c4e3759","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-19T17:30:44Z","title_canon_sha256":"5f1a0afb8b779833165ca4493be62cf743c44c2fc31df58b40c9005e76e0e73b"},"schema_version":"1.0","source":{"id":"2303.10720","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2303.10720","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"arxiv_version","alias_value":"2303.10720v2","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.10720","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"pith_short_12","alias_value":"XUFIRCUTFRYO","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"pith_short_16","alias_value":"XUFIRCUTFRYOYI3X","created_at":"2026-07-05T05:55:22Z"},{"alias_kind":"pith_short_8","alias_value":"XUFIRCUT","created_at":"2026-07-05T05:55:22Z"}],"graph_snapshots":[{"event_id":"sha256:b1b074ea3ff2f44c3e34b039d9cb1b71c842e9b451a9f66cce437e6b380f136b","target":"graph","created_at":"2026-07-05T05:55:22Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2303.10720/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent studies on transfer learning have shown that selectively fine-tuning a subset of layers or customizing different learning rates for each layer can greatly improve robustness to out-of-distribution (OOD) data and retain generalization capability in the pre-trained models. However, most of these methods employ manually crafted heuristics or expensive hyper-parameter searches, which prevent them from scaling up to large datasets and neural networks. To solve this problem, we propose Trainable Projected Gradient Method (TPGM) to automatically learn the constraint imposed for each layer for ","authors_text":"Chih-Yao Ma, Junjiao Tian, Xiaoliang Dai, Yen-Cheng Liu, Zecheng He, Zsolt Kira","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-19T17:30:44Z","title":"Trainable Projected Gradient Method for Robust Fine-tuning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.10720","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9e3d2a7bf041ad09a511cce6de0c8796b7775a4301851762d01b39ecd4bd3992","target":"record","created_at":"2026-07-05T05:55:22Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"91dbad81ec0527d32a91e8ba8a5cabe74f4a1d471d8ada37612ddd627c4e3759","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-19T17:30:44Z","title_canon_sha256":"5f1a0afb8b779833165ca4493be62cf743c44c2fc31df58b40c9005e76e0e73b"},"schema_version":"1.0","source":{"id":"2303.10720","kind":"arxiv","version":2}},"canonical_sha256":"bd0a888a932c70ec23777a6b014bf8a5e64af6801d4ac8cb66d374b4221b6dde","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bd0a888a932c70ec23777a6b014bf8a5e64af6801d4ac8cb66d374b4221b6dde","first_computed_at":"2026-07-05T05:55:22.152727Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:55:22.152727Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"dBgQqZaFRkN1rKuh//Fa4s6mPl815Eorlt4nhx/4A9rnMFiy6W/dcvV5CtOR/q5KepAtBPaLQVTGXXtR+5iGAg==","signature_status":"signed_v1","signed_at":"2026-07-05T05:55:22.153106Z","signed_message":"canonical_sha256_bytes"},"source_id":"2303.10720","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9e3d2a7bf041ad09a511cce6de0c8796b7775a4301851762d01b39ecd4bd3992","sha256:b1b074ea3ff2f44c3e34b039d9cb1b71c842e9b451a9f66cce437e6b380f136b"],"state_sha256":"65b8768fff4e8e142fda1e07a9da8c8549a4577a99416400fa405a674b1fb523"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/DfCeajSvNV4pVNkSj1/484Vf5buQhsWAiwEWVyOmTy1IhHpMHo47/oWAlBV53EihWQwMpt/0C7Q01/YewGcCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T14:39:58.604898Z","bundle_sha256":"81c26261ff32d300dadee1f1159d2fdddd82adc76d371c652da1e3cd774bdc02"}}