{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:5NXVMLVO6TVXTBDFWSM766LGNW","short_pith_number":"pith:5NXVMLVO","canonical_record":{"source":{"id":"2202.02831","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2022-02-06T18:52:21Z","cross_cats_sorted":["cs.LG","math.OC"],"title_canon_sha256":"c1a246dee3d99cc0b68da8a8d0252bd50c21cd5e12c7a0bb97d33bbdb3b04a4f","abstract_canon_sha256":"ee16f7c7d0a56b67d5a4f020008f73488e09907a5f9f61136296d66ab253d852"},"schema_version":"1.0"},"canonical_sha256":"eb6f562eaef4eb798465b499ff79666d9c1ecf1d4d110bda2f12c18863ffa3a6","source":{"kind":"arxiv","id":"2202.02831","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2202.02831","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"arxiv_version","alias_value":"2202.02831v3","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.02831","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"pith_short_12","alias_value":"5NXVMLVO6TVX","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"pith_short_16","alias_value":"5NXVMLVO6TVXTBDF","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"pith_short_8","alias_value":"5NXVMLVO","created_at":"2026-07-05T06:11:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:5NXVMLVO6TVXTBDFWSM766LGNW","target":"record","payload":{"canonical_record":{"source":{"id":"2202.02831","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2022-02-06T18:52:21Z","cross_cats_sorted":["cs.LG","math.OC"],"title_canon_sha256":"c1a246dee3d99cc0b68da8a8d0252bd50c21cd5e12c7a0bb97d33bbdb3b04a4f","abstract_canon_sha256":"ee16f7c7d0a56b67d5a4f020008f73488e09907a5f9f61136296d66ab253d852"},"schema_version":"1.0"},"canonical_sha256":"eb6f562eaef4eb798465b499ff79666d9c1ecf1d4d110bda2f12c18863ffa3a6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:11:41.800404Z","signature_b64":"gDWhPuawnRWHgb5WJeY39BlxT+aIe9JU8Qc+xPfVm9FTz43XdXPfsS+qA8FgJY/735V2+b9p1xSb3X3Nf9P4CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eb6f562eaef4eb798465b499ff79666d9c1ecf1d4d110bda2f12c18863ffa3a6","last_reissued_at":"2026-07-05T06:11:41.799897Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:11:41.799897Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2202.02831","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":"2026-07-05T06:11:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"n6k6X2VHkC7XJYGg2bIOCJiefG34B0JeZNWkw/8UhOH/eHmSgvGJ4AyAXjvfXWk1z2JREbPgUh/9DoG7k6rMDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T19:13:59.904033Z"},"content_sha256":"c814546d7ff9e3271912aac222104e31b347f5d21e23d7432c6c49b41df97c38","schema_version":"1.0","event_id":"sha256:c814546d7ff9e3271912aac222104e31b347f5d21e23d7432c6c49b41df97c38"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:5NXVMLVO6TVXTBDFWSM766LGNW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Anticorrelated Noise Injection for Improved Generalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.OC"],"primary_cat":"stat.ML","authors_text":"Antonio Orvieto, Aurelien Lucchi, Francis Bach, Frank Proske, Hans Kersting","submitted_at":"2022-02-06T18:52:21Z","abstract_excerpt":"Injecting artificial noise into gradient descent (GD) is commonly employed to improve the performance of machine learning models. Usually, uncorrelated noise is used in such perturbed gradient descent (PGD) methods. It is, however, not known if this is optimal or whether other types of noise could provide better generalization performance. In this paper, we zoom in on the problem of correlating the perturbations of consecutive PGD steps. We consider a variety of objective functions for which we find that GD with anticorrelated perturbations (\"Anti-PGD\") generalizes significantly better than GD"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.02831","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/2202.02831/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":"2026-07-05T06:11:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZlkuRXJ1dwOMNvoZilO4nkb2Jx+aRvICFNst6CXRj2WiFZiGfjTa68itkpAjKQ7Er8kc00V02akKnM0hJI7eBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T19:13:59.904730Z"},"content_sha256":"b991ab72985cf4521515c8b43a5100f5b3a5a1d8a58e4a6d8d88e5883e1fe518","schema_version":"1.0","event_id":"sha256:b991ab72985cf4521515c8b43a5100f5b3a5a1d8a58e4a6d8d88e5883e1fe518"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/5NXVMLVO6TVXTBDFWSM766LGNW/bundle.json","state_url":"https://pith.science/pith/5NXVMLVO6TVXTBDFWSM766LGNW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/5NXVMLVO6TVXTBDFWSM766LGNW/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":"2026-08-09T19:13:59Z","links":{"resolver":"https://pith.science/pith/5NXVMLVO6TVXTBDFWSM766LGNW","bundle":"https://pith.science/pith/5NXVMLVO6TVXTBDFWSM766LGNW/bundle.json","state":"https://pith.science/pith/5NXVMLVO6TVXTBDFWSM766LGNW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/5NXVMLVO6TVXTBDFWSM766LGNW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:5NXVMLVO6TVXTBDFWSM766LGNW","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":"ee16f7c7d0a56b67d5a4f020008f73488e09907a5f9f61136296d66ab253d852","cross_cats_sorted":["cs.LG","math.OC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2022-02-06T18:52:21Z","title_canon_sha256":"c1a246dee3d99cc0b68da8a8d0252bd50c21cd5e12c7a0bb97d33bbdb3b04a4f"},"schema_version":"1.0","source":{"id":"2202.02831","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2202.02831","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"arxiv_version","alias_value":"2202.02831v3","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.02831","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"pith_short_12","alias_value":"5NXVMLVO6TVX","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"pith_short_16","alias_value":"5NXVMLVO6TVXTBDF","created_at":"2026-07-05T06:11:41Z"},{"alias_kind":"pith_short_8","alias_value":"5NXVMLVO","created_at":"2026-07-05T06:11:41Z"}],"graph_snapshots":[{"event_id":"sha256:b991ab72985cf4521515c8b43a5100f5b3a5a1d8a58e4a6d8d88e5883e1fe518","target":"graph","created_at":"2026-07-05T06:11:41Z","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/2202.02831/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Injecting artificial noise into gradient descent (GD) is commonly employed to improve the performance of machine learning models. Usually, uncorrelated noise is used in such perturbed gradient descent (PGD) methods. It is, however, not known if this is optimal or whether other types of noise could provide better generalization performance. In this paper, we zoom in on the problem of correlating the perturbations of consecutive PGD steps. We consider a variety of objective functions for which we find that GD with anticorrelated perturbations (\"Anti-PGD\") generalizes significantly better than GD","authors_text":"Antonio Orvieto, Aurelien Lucchi, Francis Bach, Frank Proske, Hans Kersting","cross_cats":["cs.LG","math.OC"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2022-02-06T18:52:21Z","title":"Anticorrelated Noise Injection for Improved Generalization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.02831","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:c814546d7ff9e3271912aac222104e31b347f5d21e23d7432c6c49b41df97c38","target":"record","created_at":"2026-07-05T06:11:41Z","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":"ee16f7c7d0a56b67d5a4f020008f73488e09907a5f9f61136296d66ab253d852","cross_cats_sorted":["cs.LG","math.OC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2022-02-06T18:52:21Z","title_canon_sha256":"c1a246dee3d99cc0b68da8a8d0252bd50c21cd5e12c7a0bb97d33bbdb3b04a4f"},"schema_version":"1.0","source":{"id":"2202.02831","kind":"arxiv","version":3}},"canonical_sha256":"eb6f562eaef4eb798465b499ff79666d9c1ecf1d4d110bda2f12c18863ffa3a6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"eb6f562eaef4eb798465b499ff79666d9c1ecf1d4d110bda2f12c18863ffa3a6","first_computed_at":"2026-07-05T06:11:41.799897Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:11:41.799897Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"gDWhPuawnRWHgb5WJeY39BlxT+aIe9JU8Qc+xPfVm9FTz43XdXPfsS+qA8FgJY/735V2+b9p1xSb3X3Nf9P4CA==","signature_status":"signed_v1","signed_at":"2026-07-05T06:11:41.800404Z","signed_message":"canonical_sha256_bytes"},"source_id":"2202.02831","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c814546d7ff9e3271912aac222104e31b347f5d21e23d7432c6c49b41df97c38","sha256:b991ab72985cf4521515c8b43a5100f5b3a5a1d8a58e4a6d8d88e5883e1fe518"],"state_sha256":"330493c50cd2487de995997b96a1befe7b40f1e95076538294580e9ef22ebdde"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rwBlbsps+KY83YrJC3p0mYugZxZd45xae6oEPHTnndzPKKOzwRSLVFvUXDMqGqR04Y9Mgt8egzWAoyOPtPg/AQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T19:13:59.909683Z","bundle_sha256":"6f3c82551272ea8b50629709ed9266655e857225e4d306effda0165b0ea89b37"}}