{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:3HMKMB2TT72FSW2V2MQYZGB7RN","short_pith_number":"pith:3HMKMB2T","canonical_record":{"source":{"id":"2510.01137","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-01T17:25:23Z","cross_cats_sorted":[],"title_canon_sha256":"656c0453458b969ca79c17ab4a67c89aede925e9751251f53df9721a96d587a8","abstract_canon_sha256":"6d1cef6bbd17e8a71c1cab76ec494311cf355990ca7524cfd207d5dc0d5c7977"},"schema_version":"1.0"},"canonical_sha256":"d9d8a607539ff4595b55d3218c983f8b62638bb4964ac85ad6be3c42e405109b","source":{"kind":"arxiv","id":"2510.01137","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.01137","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"arxiv_version","alias_value":"2510.01137v3","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.01137","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"pith_short_12","alias_value":"3HMKMB2TT72F","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"pith_short_16","alias_value":"3HMKMB2TT72FSW2V","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"pith_short_8","alias_value":"3HMKMB2T","created_at":"2026-06-01T01:02:22Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:3HMKMB2TT72FSW2V2MQYZGB7RN","target":"record","payload":{"canonical_record":{"source":{"id":"2510.01137","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-01T17:25:23Z","cross_cats_sorted":[],"title_canon_sha256":"656c0453458b969ca79c17ab4a67c89aede925e9751251f53df9721a96d587a8","abstract_canon_sha256":"6d1cef6bbd17e8a71c1cab76ec494311cf355990ca7524cfd207d5dc0d5c7977"},"schema_version":"1.0"},"canonical_sha256":"d9d8a607539ff4595b55d3218c983f8b62638bb4964ac85ad6be3c42e405109b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:02:22.190299Z","signature_b64":"WqODxT95yZa1cPTWVFeF4bVkIhcHScde9MUnUQ/WZfytzxfYh5pBlM0pUnpJ57NIwdUMHlGVIgLP6vdLfws6Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d9d8a607539ff4595b55d3218c983f8b62638bb4964ac85ad6be3c42e405109b","last_reissued_at":"2026-06-01T01:02:22.188983Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:02:22.188983Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2510.01137","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-01T01:02:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aeN7XXowGcAi1IOYlzIUE+2AjzViiRIihM9zl2l3ffr91I5JdsRofah7hm8qrIgJIsHml5jEHevGt0eQSwytDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-01T23:03:24.005156Z"},"content_sha256":"ebd9d442a7b4852e398d9b0590ee93fff689705ec617f2bd254bdcc01040000b","schema_version":"1.0","event_id":"sha256:ebd9d442a7b4852e398d9b0590ee93fff689705ec617f2bd254bdcc01040000b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:3HMKMB2TT72FSW2V2MQYZGB7RN","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Re-examining Low Rank adaptation for private LLM fine-tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ali Dadsetan, Frank Rudzicz","submitted_at":"2025-10-01T17:25:23Z","abstract_excerpt":"Privacy is a central concern when fine-tuning large language models (LLMs) on sensitive data, and differentially private stochastic gradient descent (DP-SGD) -- which clips per-sample gradients and adds calibrated Gaussian noise -- is the standard tool for formal privacy guarantees. Both theory and practice show that lower-rank models are better suited to DP training, a property especially relevant for LLMs, whose fine-tuning gradients exhibit a strong low-rank structure. Methods such as DP-LoRA exploit this by restricting updates to a low-rank subspace, i.e., retaining only a few non-zero com"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.01137","kind":"arxiv","version":3},"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/2510.01137/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":"2026-06-01T01:02:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"vLor1v854FK0BSXtwvdjPC4gKMKOQJqZaz+M3dUbIgGtHcA/Fi05jGg3VpaBZEGBU35D/BSQ2FCdzU225TWTAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-01T23:03:24.005550Z"},"content_sha256":"2721f4c49f90fb091ab612b4400a5e451724975f17a97e8c17994acdfbecda43","schema_version":"1.0","event_id":"sha256:2721f4c49f90fb091ab612b4400a5e451724975f17a97e8c17994acdfbecda43"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3HMKMB2TT72FSW2V2MQYZGB7RN/bundle.json","state_url":"https://pith.science/pith/3HMKMB2TT72FSW2V2MQYZGB7RN/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3HMKMB2TT72FSW2V2MQYZGB7RN/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":"2026-06-01T23:03:24Z","links":{"resolver":"https://pith.science/pith/3HMKMB2TT72FSW2V2MQYZGB7RN","bundle":"https://pith.science/pith/3HMKMB2TT72FSW2V2MQYZGB7RN/bundle.json","state":"https://pith.science/pith/3HMKMB2TT72FSW2V2MQYZGB7RN/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3HMKMB2TT72FSW2V2MQYZGB7RN/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3HMKMB2TT72FSW2V2MQYZGB7RN","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":"6d1cef6bbd17e8a71c1cab76ec494311cf355990ca7524cfd207d5dc0d5c7977","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-01T17:25:23Z","title_canon_sha256":"656c0453458b969ca79c17ab4a67c89aede925e9751251f53df9721a96d587a8"},"schema_version":"1.0","source":{"id":"2510.01137","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.01137","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"arxiv_version","alias_value":"2510.01137v3","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.01137","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"pith_short_12","alias_value":"3HMKMB2TT72F","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"pith_short_16","alias_value":"3HMKMB2TT72FSW2V","created_at":"2026-06-01T01:02:22Z"},{"alias_kind":"pith_short_8","alias_value":"3HMKMB2T","created_at":"2026-06-01T01:02:22Z"}],"graph_snapshots":[{"event_id":"sha256:2721f4c49f90fb091ab612b4400a5e451724975f17a97e8c17994acdfbecda43","target":"graph","created_at":"2026-06-01T01:02:22Z","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/2510.01137/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Privacy is a central concern when fine-tuning large language models (LLMs) on sensitive data, and differentially private stochastic gradient descent (DP-SGD) -- which clips per-sample gradients and adds calibrated Gaussian noise -- is the standard tool for formal privacy guarantees. Both theory and practice show that lower-rank models are better suited to DP training, a property especially relevant for LLMs, whose fine-tuning gradients exhibit a strong low-rank structure. Methods such as DP-LoRA exploit this by restricting updates to a low-rank subspace, i.e., retaining only a few non-zero com","authors_text":"Ali Dadsetan, Frank Rudzicz","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-01T17:25:23Z","title":"Re-examining Low Rank adaptation for private LLM fine-tuning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.01137","kind":"arxiv","version":3},"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:ebd9d442a7b4852e398d9b0590ee93fff689705ec617f2bd254bdcc01040000b","target":"record","created_at":"2026-06-01T01:02:22Z","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":"6d1cef6bbd17e8a71c1cab76ec494311cf355990ca7524cfd207d5dc0d5c7977","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-01T17:25:23Z","title_canon_sha256":"656c0453458b969ca79c17ab4a67c89aede925e9751251f53df9721a96d587a8"},"schema_version":"1.0","source":{"id":"2510.01137","kind":"arxiv","version":3}},"canonical_sha256":"d9d8a607539ff4595b55d3218c983f8b62638bb4964ac85ad6be3c42e405109b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d9d8a607539ff4595b55d3218c983f8b62638bb4964ac85ad6be3c42e405109b","first_computed_at":"2026-06-01T01:02:22.188983Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-01T01:02:22.188983Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"WqODxT95yZa1cPTWVFeF4bVkIhcHScde9MUnUQ/WZfytzxfYh5pBlM0pUnpJ57NIwdUMHlGVIgLP6vdLfws6Bg==","signature_status":"signed_v1","signed_at":"2026-06-01T01:02:22.190299Z","signed_message":"canonical_sha256_bytes"},"source_id":"2510.01137","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ebd9d442a7b4852e398d9b0590ee93fff689705ec617f2bd254bdcc01040000b","sha256:2721f4c49f90fb091ab612b4400a5e451724975f17a97e8c17994acdfbecda43"],"state_sha256":"d6b43c80352d5a751357f628ebbcc24f15e34895b47c9a5fd4f7123c10a77381"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7qEzZtlUC0+tvXCTcCQhoxJ9nTfuLwKQru3P/JOaPMyQ9qxP5ai3f4zkmDG2Jx9Fu7mpZkcdlO8YIonAYZPWDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-01T23:03:24.007566Z","bundle_sha256":"a28b0c08072b20f94b193322cf2b96c66bfae1aecc488479930812b649a55074"}}