{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:FIOAEJH3OPDQXHHOT3FCXXF6HL","short_pith_number":"pith:FIOAEJH3","canonical_record":{"source":{"id":"2302.06598","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-13T18:54:58Z","cross_cats_sorted":[],"title_canon_sha256":"7b3013e3de47ba93bd855ec86393c7c289ce11007d8d46aabf1a7d5604e6fd43","abstract_canon_sha256":"404c9a15cd46e00157d3bbc455217ccbf11dc3922597b13d2b7f995d9a61d2dd"},"schema_version":"1.0"},"canonical_sha256":"2a1c0224fb73c70b9cee9eca2bdcbe3af53199e83e2f91a5c673aca99594032c","source":{"kind":"arxiv","id":"2302.06598","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2302.06598","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"arxiv_version","alias_value":"2302.06598v1","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.06598","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"pith_short_12","alias_value":"FIOAEJH3OPDQ","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"pith_short_16","alias_value":"FIOAEJH3OPDQXHHO","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"pith_short_8","alias_value":"FIOAEJH3","created_at":"2026-07-05T05:41:12Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:FIOAEJH3OPDQXHHOT3FCXXF6HL","target":"record","payload":{"canonical_record":{"source":{"id":"2302.06598","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-13T18:54:58Z","cross_cats_sorted":[],"title_canon_sha256":"7b3013e3de47ba93bd855ec86393c7c289ce11007d8d46aabf1a7d5604e6fd43","abstract_canon_sha256":"404c9a15cd46e00157d3bbc455217ccbf11dc3922597b13d2b7f995d9a61d2dd"},"schema_version":"1.0"},"canonical_sha256":"2a1c0224fb73c70b9cee9eca2bdcbe3af53199e83e2f91a5c673aca99594032c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:41:12.745916Z","signature_b64":"byuirr+pu8bgR9jwqR5mzARVxkGnOcDevCxeQ//wn/MK8HD4TMTEsXQGR13VnBe+QgMtO//04NMlCEBzMHV/Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2a1c0224fb73c70b9cee9eca2bdcbe3af53199e83e2f91a5c673aca99594032c","last_reissued_at":"2026-07-05T05:41:12.745481Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:41:12.745481Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2302.06598","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:41:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qsI27xbhYPu3jF/6eH6QoPuH8JvLhIpgDgg3s7y9LXOo40j7ik4IbrlcPkXD3OgWc9p8XJ9VIMe1gxHf+UzsCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:42:03.077351Z"},"content_sha256":"777d783ef5036004daab45eae14b8baa79601132360e1836b8132ee357c3732c","schema_version":"1.0","event_id":"sha256:777d783ef5036004daab45eae14b8baa79601132360e1836b8132ee357c3732c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:FIOAEJH3OPDQXHHOT3FCXXF6HL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Gradient-Based Automated Iterative Recovery for Parameter-Efficient Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ann Yuan, Frederick Liu, Lucas Dixon, Maximilian Mozes, Nithum Thain, Tolga Bolukbasi","submitted_at":"2023-02-13T18:54:58Z","abstract_excerpt":"Pretrained large language models (LLMs) are able to solve a wide variety of tasks through transfer learning. Various explainability methods have been developed to investigate their decision making process. TracIn (Pruthi et al., 2020) is one such gradient-based method which explains model inferences based on the influence of training examples. In this paper, we explore the use of TracIn to improve model performance in the parameter-efficient tuning (PET) setting. We develop conversational safety classifiers via the prompt-tuning PET method and show how the unique characteristics of the PET reg"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.06598","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/2302.06598/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":"2026-07-05T05:41:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GCJsLAZzAIIkbslIi72ANgAw4zKSwO7ngdutrdaXEOmsJLWU4ie9ibDS6LBUMP7ELP7Wsia4tR2ubvVojM5RBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:42:03.077919Z"},"content_sha256":"8dce0b6fc0761f7757aaa966dab64c73be82cf9451ac44b014a76bd4bd619b22","schema_version":"1.0","event_id":"sha256:8dce0b6fc0761f7757aaa966dab64c73be82cf9451ac44b014a76bd4bd619b22"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FIOAEJH3OPDQXHHOT3FCXXF6HL/bundle.json","state_url":"https://pith.science/pith/FIOAEJH3OPDQXHHOT3FCXXF6HL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FIOAEJH3OPDQXHHOT3FCXXF6HL/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":"2026-08-09T00:42:03Z","links":{"resolver":"https://pith.science/pith/FIOAEJH3OPDQXHHOT3FCXXF6HL","bundle":"https://pith.science/pith/FIOAEJH3OPDQXHHOT3FCXXF6HL/bundle.json","state":"https://pith.science/pith/FIOAEJH3OPDQXHHOT3FCXXF6HL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FIOAEJH3OPDQXHHOT3FCXXF6HL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:FIOAEJH3OPDQXHHOT3FCXXF6HL","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":"404c9a15cd46e00157d3bbc455217ccbf11dc3922597b13d2b7f995d9a61d2dd","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-13T18:54:58Z","title_canon_sha256":"7b3013e3de47ba93bd855ec86393c7c289ce11007d8d46aabf1a7d5604e6fd43"},"schema_version":"1.0","source":{"id":"2302.06598","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2302.06598","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"arxiv_version","alias_value":"2302.06598v1","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.06598","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"pith_short_12","alias_value":"FIOAEJH3OPDQ","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"pith_short_16","alias_value":"FIOAEJH3OPDQXHHO","created_at":"2026-07-05T05:41:12Z"},{"alias_kind":"pith_short_8","alias_value":"FIOAEJH3","created_at":"2026-07-05T05:41:12Z"}],"graph_snapshots":[{"event_id":"sha256:8dce0b6fc0761f7757aaa966dab64c73be82cf9451ac44b014a76bd4bd619b22","target":"graph","created_at":"2026-07-05T05:41:12Z","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/2302.06598/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Pretrained large language models (LLMs) are able to solve a wide variety of tasks through transfer learning. Various explainability methods have been developed to investigate their decision making process. TracIn (Pruthi et al., 2020) is one such gradient-based method which explains model inferences based on the influence of training examples. In this paper, we explore the use of TracIn to improve model performance in the parameter-efficient tuning (PET) setting. We develop conversational safety classifiers via the prompt-tuning PET method and show how the unique characteristics of the PET reg","authors_text":"Ann Yuan, Frederick Liu, Lucas Dixon, Maximilian Mozes, Nithum Thain, Tolga Bolukbasi","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-13T18:54:58Z","title":"Gradient-Based Automated Iterative Recovery for Parameter-Efficient Tuning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.06598","kind":"arxiv","version":1},"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":"sha256:777d783ef5036004daab45eae14b8baa79601132360e1836b8132ee357c3732c","target":"record","created_at":"2026-07-05T05:41:12Z","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":"404c9a15cd46e00157d3bbc455217ccbf11dc3922597b13d2b7f995d9a61d2dd","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-13T18:54:58Z","title_canon_sha256":"7b3013e3de47ba93bd855ec86393c7c289ce11007d8d46aabf1a7d5604e6fd43"},"schema_version":"1.0","source":{"id":"2302.06598","kind":"arxiv","version":1}},"canonical_sha256":"2a1c0224fb73c70b9cee9eca2bdcbe3af53199e83e2f91a5c673aca99594032c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2a1c0224fb73c70b9cee9eca2bdcbe3af53199e83e2f91a5c673aca99594032c","first_computed_at":"2026-07-05T05:41:12.745481Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:41:12.745481Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"byuirr+pu8bgR9jwqR5mzARVxkGnOcDevCxeQ//wn/MK8HD4TMTEsXQGR13VnBe+QgMtO//04NMlCEBzMHV/Dw==","signature_status":"signed_v1","signed_at":"2026-07-05T05:41:12.745916Z","signed_message":"canonical_sha256_bytes"},"source_id":"2302.06598","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:777d783ef5036004daab45eae14b8baa79601132360e1836b8132ee357c3732c","sha256:8dce0b6fc0761f7757aaa966dab64c73be82cf9451ac44b014a76bd4bd619b22"],"state_sha256":"7bf72a1792ce15f5f3fe969807258302926a55d95c734bb46186c84e58d377f0"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yhPAI7ymQq/xFG3Y9PpkO7txftVhlRauIC/AOh31G4pmbBtc/BAFuhViH2emTS1c44GcBMuq5cp11Jy8I/7xBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T00:42:03.082502Z","bundle_sha256":"7fe16bb1697c6313c170d6f6416f6c5bee90203e2ee9d6adff009864d8f11cb5"}}