{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:FNRMSRHDI6PGOMEGKRGDWJMW3M","short_pith_number":"pith:FNRMSRHD","canonical_record":{"source":{"id":"2201.12328","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-01-28T18:48:18Z","cross_cats_sorted":[],"title_canon_sha256":"ce7bf3bdfd1a18076dfab28a8f4d65ff98e4d5087dba3ef0fc4ef1f072c14df0","abstract_canon_sha256":"c4c0b8b9dc790f8850c525e3a01dce698f2dea4fb6d61a4f73540acae5c2bfd5"},"schema_version":"1.0"},"canonical_sha256":"2b62c944e3479e673086544c3b2596db261eb4411d6f02ce50d80f85ff2b4baa","source":{"kind":"arxiv","id":"2201.12328","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2201.12328","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"arxiv_version","alias_value":"2201.12328v2","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2201.12328","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"pith_short_12","alias_value":"FNRMSRHDI6PG","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"pith_short_16","alias_value":"FNRMSRHDI6PGOMEG","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"pith_short_8","alias_value":"FNRMSRHD","created_at":"2026-07-05T03:55:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:FNRMSRHDI6PGOMEGKRGDWJMW3M","target":"record","payload":{"canonical_record":{"source":{"id":"2201.12328","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-01-28T18:48:18Z","cross_cats_sorted":[],"title_canon_sha256":"ce7bf3bdfd1a18076dfab28a8f4d65ff98e4d5087dba3ef0fc4ef1f072c14df0","abstract_canon_sha256":"c4c0b8b9dc790f8850c525e3a01dce698f2dea4fb6d61a4f73540acae5c2bfd5"},"schema_version":"1.0"},"canonical_sha256":"2b62c944e3479e673086544c3b2596db261eb4411d6f02ce50d80f85ff2b4baa","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:55:32.213029Z","signature_b64":"pcY4beZEN1J/qLY2JqlEDF9FYHFmyqJvcVvHVmAyrOwSIZOLdHJpBACer7BwE2m9cykUyfFf+TS4qrA/0vyVAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2b62c944e3479e673086544c3b2596db261eb4411d6f02ce50d80f85ff2b4baa","last_reissued_at":"2026-07-05T03:55:32.212451Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:55:32.212451Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2201.12328","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":"2026-07-05T03:55:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/6TQf+yT9Fm1glrbhVMYJhvkLaWkD1NCRVHTgUPPCkRvvWoDYK16dAik6CK7slqEBA8ibn0CVrRArItY1pqADw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T13:20:11.720682Z"},"content_sha256":"2660bafc3519f902702951fffeac77b280df7fc459e2a481216cc2e38513e573","schema_version":"1.0","event_id":"sha256:2660bafc3519f902702951fffeac77b280df7fc459e2a481216cc2e38513e573"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:FNRMSRHDI6PGOMEGKRGDWJMW3M","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Toward Training at ImageNet Scale with Differential Privacy","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Abhradeep Thakurta, Alexey Kurakin, Andreas Terzis, Roxana Geambasu, Shuang Song, Steve Chien","submitted_at":"2022-01-28T18:48:18Z","abstract_excerpt":"Differential privacy (DP) is the de facto standard for training machine learning (ML) models, including neural networks, while ensuring the privacy of individual examples in the training set. Despite a rich literature on how to train ML models with differential privacy, it remains extremely challenging to train real-life, large neural networks with both reasonable accuracy and privacy.\n  We set out to investigate how to do this, using ImageNet image classification as a poster example of an ML task that is very challenging to resolve accurately with DP right now. This paper shares initial lesso"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2201.12328","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/2201.12328/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":"2026-07-05T03:55:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TV7aF2OCjwf0TsfLE23KWeXQ/w1IS4o5F2tYXEfMpkml8zCjx13iCX2JZl0kN1H9Ul3HWxx7ZdzoFh5lyCaMDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T13:20:11.721234Z"},"content_sha256":"92f37a3da1ac6d14aa7ae5e6a98d6d5563b4613c28f404c447ba39d522dd4171","schema_version":"1.0","event_id":"sha256:92f37a3da1ac6d14aa7ae5e6a98d6d5563b4613c28f404c447ba39d522dd4171"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FNRMSRHDI6PGOMEGKRGDWJMW3M/bundle.json","state_url":"https://pith.science/pith/FNRMSRHDI6PGOMEGKRGDWJMW3M/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FNRMSRHDI6PGOMEGKRGDWJMW3M/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":"2026-08-04T13:20:11Z","links":{"resolver":"https://pith.science/pith/FNRMSRHDI6PGOMEGKRGDWJMW3M","bundle":"https://pith.science/pith/FNRMSRHDI6PGOMEGKRGDWJMW3M/bundle.json","state":"https://pith.science/pith/FNRMSRHDI6PGOMEGKRGDWJMW3M/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FNRMSRHDI6PGOMEGKRGDWJMW3M/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:FNRMSRHDI6PGOMEGKRGDWJMW3M","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":"c4c0b8b9dc790f8850c525e3a01dce698f2dea4fb6d61a4f73540acae5c2bfd5","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-01-28T18:48:18Z","title_canon_sha256":"ce7bf3bdfd1a18076dfab28a8f4d65ff98e4d5087dba3ef0fc4ef1f072c14df0"},"schema_version":"1.0","source":{"id":"2201.12328","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2201.12328","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"arxiv_version","alias_value":"2201.12328v2","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2201.12328","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"pith_short_12","alias_value":"FNRMSRHDI6PG","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"pith_short_16","alias_value":"FNRMSRHDI6PGOMEG","created_at":"2026-07-05T03:55:32Z"},{"alias_kind":"pith_short_8","alias_value":"FNRMSRHD","created_at":"2026-07-05T03:55:32Z"}],"graph_snapshots":[{"event_id":"sha256:92f37a3da1ac6d14aa7ae5e6a98d6d5563b4613c28f404c447ba39d522dd4171","target":"graph","created_at":"2026-07-05T03:55:32Z","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/2201.12328/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Differential privacy (DP) is the de facto standard for training machine learning (ML) models, including neural networks, while ensuring the privacy of individual examples in the training set. Despite a rich literature on how to train ML models with differential privacy, it remains extremely challenging to train real-life, large neural networks with both reasonable accuracy and privacy.\n  We set out to investigate how to do this, using ImageNet image classification as a poster example of an ML task that is very challenging to resolve accurately with DP right now. This paper shares initial lesso","authors_text":"Abhradeep Thakurta, Alexey Kurakin, Andreas Terzis, Roxana Geambasu, Shuang Song, Steve Chien","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-01-28T18:48:18Z","title":"Toward Training at ImageNet Scale with Differential Privacy"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2201.12328","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":"sha256:2660bafc3519f902702951fffeac77b280df7fc459e2a481216cc2e38513e573","target":"record","created_at":"2026-07-05T03:55:32Z","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":"c4c0b8b9dc790f8850c525e3a01dce698f2dea4fb6d61a4f73540acae5c2bfd5","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-01-28T18:48:18Z","title_canon_sha256":"ce7bf3bdfd1a18076dfab28a8f4d65ff98e4d5087dba3ef0fc4ef1f072c14df0"},"schema_version":"1.0","source":{"id":"2201.12328","kind":"arxiv","version":2}},"canonical_sha256":"2b62c944e3479e673086544c3b2596db261eb4411d6f02ce50d80f85ff2b4baa","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2b62c944e3479e673086544c3b2596db261eb4411d6f02ce50d80f85ff2b4baa","first_computed_at":"2026-07-05T03:55:32.212451Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T03:55:32.212451Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"pcY4beZEN1J/qLY2JqlEDF9FYHFmyqJvcVvHVmAyrOwSIZOLdHJpBACer7BwE2m9cykUyfFf+TS4qrA/0vyVAA==","signature_status":"signed_v1","signed_at":"2026-07-05T03:55:32.213029Z","signed_message":"canonical_sha256_bytes"},"source_id":"2201.12328","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2660bafc3519f902702951fffeac77b280df7fc459e2a481216cc2e38513e573","sha256:92f37a3da1ac6d14aa7ae5e6a98d6d5563b4613c28f404c447ba39d522dd4171"],"state_sha256":"dad1141834cd852eb1ec21c7358d2b129bc4533d1c00ed71d193bc9201e9cc94"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Qp/1PUF6ICCx1bB2sTzLcGznVjkYAvauT9ARmoIbo1db0boWolZW1J+yljQLKb9zF3riWYNUDcSQgvGW7wyABw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T13:20:11.725180Z","bundle_sha256":"bfea779c5942522059a81477d41d8a173ee9b9a8a7ccf8ebf7e6fa60a38b7fba"}}