{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:MWSCWHWQFN3JZUDEJ3TK6FVTEQ","short_pith_number":"pith:MWSCWHWQ","canonical_record":{"source":{"id":"2411.15831","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-24T13:17:36Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6abdf3428e55281956ec0b9a31d3022be269c943f419b66f805c81a99112aeab","abstract_canon_sha256":"bef9fe14948bfc1318248eaaf5f17748d66663f0cf29207a6f6b8b0337727817"},"schema_version":"1.0"},"canonical_sha256":"65a42b1ed02b769cd0644ee6af16b3243528c5ae69d6b3769a66b24c59847b66","source":{"kind":"arxiv","id":"2411.15831","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.15831","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"arxiv_version","alias_value":"2411.15831v1","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.15831","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"pith_short_12","alias_value":"MWSCWHWQFN3J","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"pith_short_16","alias_value":"MWSCWHWQFN3JZUDE","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"pith_short_8","alias_value":"MWSCWHWQ","created_at":"2026-07-05T09:39:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:MWSCWHWQFN3JZUDEJ3TK6FVTEQ","target":"record","payload":{"canonical_record":{"source":{"id":"2411.15831","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-24T13:17:36Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6abdf3428e55281956ec0b9a31d3022be269c943f419b66f805c81a99112aeab","abstract_canon_sha256":"bef9fe14948bfc1318248eaaf5f17748d66663f0cf29207a6f6b8b0337727817"},"schema_version":"1.0"},"canonical_sha256":"65a42b1ed02b769cd0644ee6af16b3243528c5ae69d6b3769a66b24c59847b66","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:39:41.625413Z","signature_b64":"SbIW1TYBEM6g1X2GomJXUKzc3lKkMum04cZ9SVj/IVppMf01uZapX5JwJcK9AS/c0l5SOuQRhD76AiIxqULtBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"65a42b1ed02b769cd0644ee6af16b3243528c5ae69d6b3769a66b24c59847b66","last_reissued_at":"2026-07-05T09:39:41.624019Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:39:41.624019Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2411.15831","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":"2026-07-05T09:39:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oQUDHbXgy4mRAvq1v/nj5bTzCK2dlb7tQHk901VSeuVzV/4iZZyqx85Ew+uVhkk4xhkOu+iWQVNaBI0VOp8ZBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T04:04:55.801907Z"},"content_sha256":"988c30d8960254de8ba7e70d26305b85a9ae85542fbecde63d6f537f4a177c0c","schema_version":"1.0","event_id":"sha256:988c30d8960254de8ba7e70d26305b85a9ae85542fbecde63d6f537f4a177c0c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:MWSCWHWQFN3JZUDEJ3TK6FVTEQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Efficient and Private: Memorisation under differentially private parameter-efficient fine-tuning in language models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Dmitrii Usynin, Jonathan Passerat-Palmbach, Olivia Ma","submitted_at":"2024-11-24T13:17:36Z","abstract_excerpt":"Fine-tuning large language models (LLMs) for specific tasks introduces privacy risks, as models may inadvertently memorise and leak sensitive training data. While Differential Privacy (DP) offers a solution to mitigate these risks, it introduces significant computational and performance trade-offs, particularly with standard fine-tuning approaches. Previous work has primarily focused on full-parameter updates, which are computationally intensive and may not fully leverage DPs potential in large models. In this work, we address these shortcomings by investigating Parameter-Efficient Fine-Tuning"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.15831","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/2411.15831/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":"2026-07-05T09:39:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"313n19jdt/TgIITamEkH2Gi/goSuCvAFxUaIbsOEX6fMe+IGYgvtuOMdFLunZZw5TPfkF/sM+UniNZSL9oBHAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T04:04:55.802732Z"},"content_sha256":"9b01f5e26bae2fd4893cf34c519bc5bbcf547973999812bae9e523d8c7546fe6","schema_version":"1.0","event_id":"sha256:9b01f5e26bae2fd4893cf34c519bc5bbcf547973999812bae9e523d8c7546fe6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MWSCWHWQFN3JZUDEJ3TK6FVTEQ/bundle.json","state_url":"https://pith.science/pith/MWSCWHWQFN3JZUDEJ3TK6FVTEQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MWSCWHWQFN3JZUDEJ3TK6FVTEQ/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":"2026-08-13T04:04:55Z","links":{"resolver":"https://pith.science/pith/MWSCWHWQFN3JZUDEJ3TK6FVTEQ","bundle":"https://pith.science/pith/MWSCWHWQFN3JZUDEJ3TK6FVTEQ/bundle.json","state":"https://pith.science/pith/MWSCWHWQFN3JZUDEJ3TK6FVTEQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MWSCWHWQFN3JZUDEJ3TK6FVTEQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:MWSCWHWQFN3JZUDEJ3TK6FVTEQ","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":"bef9fe14948bfc1318248eaaf5f17748d66663f0cf29207a6f6b8b0337727817","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-24T13:17:36Z","title_canon_sha256":"6abdf3428e55281956ec0b9a31d3022be269c943f419b66f805c81a99112aeab"},"schema_version":"1.0","source":{"id":"2411.15831","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.15831","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"arxiv_version","alias_value":"2411.15831v1","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.15831","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"pith_short_12","alias_value":"MWSCWHWQFN3J","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"pith_short_16","alias_value":"MWSCWHWQFN3JZUDE","created_at":"2026-07-05T09:39:41Z"},{"alias_kind":"pith_short_8","alias_value":"MWSCWHWQ","created_at":"2026-07-05T09:39:41Z"}],"graph_snapshots":[{"event_id":"sha256:9b01f5e26bae2fd4893cf34c519bc5bbcf547973999812bae9e523d8c7546fe6","target":"graph","created_at":"2026-07-05T09:39:41Z","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/2411.15831/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Fine-tuning large language models (LLMs) for specific tasks introduces privacy risks, as models may inadvertently memorise and leak sensitive training data. While Differential Privacy (DP) offers a solution to mitigate these risks, it introduces significant computational and performance trade-offs, particularly with standard fine-tuning approaches. Previous work has primarily focused on full-parameter updates, which are computationally intensive and may not fully leverage DPs potential in large models. In this work, we address these shortcomings by investigating Parameter-Efficient Fine-Tuning","authors_text":"Dmitrii Usynin, Jonathan Passerat-Palmbach, Olivia Ma","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-24T13:17:36Z","title":"Efficient and Private: Memorisation under differentially private parameter-efficient fine-tuning in language models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.15831","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":"sha256:988c30d8960254de8ba7e70d26305b85a9ae85542fbecde63d6f537f4a177c0c","target":"record","created_at":"2026-07-05T09:39:41Z","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":"bef9fe14948bfc1318248eaaf5f17748d66663f0cf29207a6f6b8b0337727817","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-24T13:17:36Z","title_canon_sha256":"6abdf3428e55281956ec0b9a31d3022be269c943f419b66f805c81a99112aeab"},"schema_version":"1.0","source":{"id":"2411.15831","kind":"arxiv","version":1}},"canonical_sha256":"65a42b1ed02b769cd0644ee6af16b3243528c5ae69d6b3769a66b24c59847b66","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"65a42b1ed02b769cd0644ee6af16b3243528c5ae69d6b3769a66b24c59847b66","first_computed_at":"2026-07-05T09:39:41.624019Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:39:41.624019Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"SbIW1TYBEM6g1X2GomJXUKzc3lKkMum04cZ9SVj/IVppMf01uZapX5JwJcK9AS/c0l5SOuQRhD76AiIxqULtBg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:39:41.625413Z","signed_message":"canonical_sha256_bytes"},"source_id":"2411.15831","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:988c30d8960254de8ba7e70d26305b85a9ae85542fbecde63d6f537f4a177c0c","sha256:9b01f5e26bae2fd4893cf34c519bc5bbcf547973999812bae9e523d8c7546fe6"],"state_sha256":"257bdda2f14f8a68ffc4699d81854491b7e640a7a8b4f8bb94861f88a9bb9322"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/4KQkRPQPW45xXiy0r362m6GL6WW7ejrRqVptIQbRauevtuPVPE70JdrtF+ZNyVOqxsIs+k7jrjWrZhrgBiiCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-13T04:04:55.812613Z","bundle_sha256":"8dd6428721f6335c3edcf7d07bfccc607ccd0d82fc3773f1e4da8e0fa43f0a92"}}