{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:4ZALQAHJ6MS3W7GNOGP5ZYCRH7","short_pith_number":"pith:4ZALQAHJ","canonical_record":{"source":{"id":"1901.09149","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T03:12:14Z","cross_cats_sorted":["math.OC","stat.ML"],"title_canon_sha256":"10a6c0db9bcbb7fa04b899f99b117f97c55a4c27a1e3a09bb15e8a9d236f7a5e","abstract_canon_sha256":"13d3ea94ed91ca6a6fa07f9ac831e3699a27bb2b78f18db5da4c1a903572844a"},"schema_version":"1.0"},"canonical_sha256":"e640b800e9f325bb7ccd719fdce0513ff81a50cb94e82c791e5b48ece137f564","source":{"kind":"arxiv","id":"1901.09149","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1901.09149","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"arxiv_version","alias_value":"1901.09149v2","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1901.09149","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"pith_short_12","alias_value":"4ZALQAHJ6MS3","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"pith_short_16","alias_value":"4ZALQAHJ6MS3W7GN","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"pith_short_8","alias_value":"4ZALQAHJ","created_at":"2026-07-05T00:37:43Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:4ZALQAHJ6MS3W7GNOGP5ZYCRH7","target":"record","payload":{"canonical_record":{"source":{"id":"1901.09149","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T03:12:14Z","cross_cats_sorted":["math.OC","stat.ML"],"title_canon_sha256":"10a6c0db9bcbb7fa04b899f99b117f97c55a4c27a1e3a09bb15e8a9d236f7a5e","abstract_canon_sha256":"13d3ea94ed91ca6a6fa07f9ac831e3699a27bb2b78f18db5da4c1a903572844a"},"schema_version":"1.0"},"canonical_sha256":"e640b800e9f325bb7ccd719fdce0513ff81a50cb94e82c791e5b48ece137f564","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:37:43.999098Z","signature_b64":"o5sWK2n0drzMz8cN8bZ23RY0cmo7vuhQLNIeWX5A/C9mem1H0r1nHNKUa0+YeC49AL8XDkls1DvQfjOQISPVBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e640b800e9f325bb7ccd719fdce0513ff81a50cb94e82c791e5b48ece137f564","last_reissued_at":"2026-07-05T00:37:43.998704Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:37:43.998704Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1901.09149","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":"2026-07-05T00:37:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"37VnngHZCE9c+swD8FvN8skSIkbZv+vDyInOPwwF8wYr/FjtohXmEVHUSZ3amASJfSNxZ+YbW9IpxYRRQVzTBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T05:55:23.831449Z"},"content_sha256":"4fc5f1205b4cba1128de06a306e525533d6fc15542aa75120310bed270bf9062","schema_version":"1.0","event_id":"sha256:4fc5f1205b4cba1128de06a306e525533d6fc15542aa75120310bed270bf9062"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:4ZALQAHJ6MS3W7GNOGP5ZYCRH7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Escaping Saddle Points with Adaptive Gradient Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Matthew Staib, Sanjiv Kumar, Sashank J. Reddi, Satyen Kale, Suvrit Sra","submitted_at":"2019-01-26T03:12:14Z","abstract_excerpt":"Adaptive methods such as Adam and RMSProp are widely used in deep learning but are not well understood. In this paper, we seek a crisp, clean and precise characterization of their behavior in nonconvex settings. To this end, we first provide a novel view of adaptive methods as preconditioned SGD, where the preconditioner is estimated in an online manner. By studying the preconditioner on its own, we elucidate its purpose: it rescales the stochastic gradient noise to be isotropic near stationary points, which helps escape saddle points. Furthermore, we show that adaptive methods can efficiently"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1901.09149","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/1901.09149/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":"2026-07-05T00:37:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aLFPsTNejA8f5kbpKu7VIIItdWRp4XwEyr7ZLBACWt42+dJKjmC7liv65O8/pqzdE22QyoI4BNI9g6HLCGwpBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T05:55:23.831964Z"},"content_sha256":"b52b0a480e223c205f2375b4bb1d7e91df1ea9ce26fce8910e8277fda09f7c32","schema_version":"1.0","event_id":"sha256:b52b0a480e223c205f2375b4bb1d7e91df1ea9ce26fce8910e8277fda09f7c32"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/4ZALQAHJ6MS3W7GNOGP5ZYCRH7/bundle.json","state_url":"https://pith.science/pith/4ZALQAHJ6MS3W7GNOGP5ZYCRH7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/4ZALQAHJ6MS3W7GNOGP5ZYCRH7/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":"2026-08-06T05:55:23Z","links":{"resolver":"https://pith.science/pith/4ZALQAHJ6MS3W7GNOGP5ZYCRH7","bundle":"https://pith.science/pith/4ZALQAHJ6MS3W7GNOGP5ZYCRH7/bundle.json","state":"https://pith.science/pith/4ZALQAHJ6MS3W7GNOGP5ZYCRH7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/4ZALQAHJ6MS3W7GNOGP5ZYCRH7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:4ZALQAHJ6MS3W7GNOGP5ZYCRH7","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":"13d3ea94ed91ca6a6fa07f9ac831e3699a27bb2b78f18db5da4c1a903572844a","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T03:12:14Z","title_canon_sha256":"10a6c0db9bcbb7fa04b899f99b117f97c55a4c27a1e3a09bb15e8a9d236f7a5e"},"schema_version":"1.0","source":{"id":"1901.09149","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1901.09149","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"arxiv_version","alias_value":"1901.09149v2","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1901.09149","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"pith_short_12","alias_value":"4ZALQAHJ6MS3","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"pith_short_16","alias_value":"4ZALQAHJ6MS3W7GN","created_at":"2026-07-05T00:37:43Z"},{"alias_kind":"pith_short_8","alias_value":"4ZALQAHJ","created_at":"2026-07-05T00:37:43Z"}],"graph_snapshots":[{"event_id":"sha256:b52b0a480e223c205f2375b4bb1d7e91df1ea9ce26fce8910e8277fda09f7c32","target":"graph","created_at":"2026-07-05T00:37:43Z","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/1901.09149/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Adaptive methods such as Adam and RMSProp are widely used in deep learning but are not well understood. In this paper, we seek a crisp, clean and precise characterization of their behavior in nonconvex settings. To this end, we first provide a novel view of adaptive methods as preconditioned SGD, where the preconditioner is estimated in an online manner. By studying the preconditioner on its own, we elucidate its purpose: it rescales the stochastic gradient noise to be isotropic near stationary points, which helps escape saddle points. Furthermore, we show that adaptive methods can efficiently","authors_text":"Matthew Staib, Sanjiv Kumar, Sashank J. Reddi, Satyen Kale, Suvrit Sra","cross_cats":["math.OC","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T03:12:14Z","title":"Escaping Saddle Points with Adaptive Gradient Methods"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1901.09149","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":"sha256:4fc5f1205b4cba1128de06a306e525533d6fc15542aa75120310bed270bf9062","target":"record","created_at":"2026-07-05T00:37:43Z","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":"13d3ea94ed91ca6a6fa07f9ac831e3699a27bb2b78f18db5da4c1a903572844a","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T03:12:14Z","title_canon_sha256":"10a6c0db9bcbb7fa04b899f99b117f97c55a4c27a1e3a09bb15e8a9d236f7a5e"},"schema_version":"1.0","source":{"id":"1901.09149","kind":"arxiv","version":2}},"canonical_sha256":"e640b800e9f325bb7ccd719fdce0513ff81a50cb94e82c791e5b48ece137f564","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e640b800e9f325bb7ccd719fdce0513ff81a50cb94e82c791e5b48ece137f564","first_computed_at":"2026-07-05T00:37:43.998704Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T00:37:43.998704Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"o5sWK2n0drzMz8cN8bZ23RY0cmo7vuhQLNIeWX5A/C9mem1H0r1nHNKUa0+YeC49AL8XDkls1DvQfjOQISPVBg==","signature_status":"signed_v1","signed_at":"2026-07-05T00:37:43.999098Z","signed_message":"canonical_sha256_bytes"},"source_id":"1901.09149","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4fc5f1205b4cba1128de06a306e525533d6fc15542aa75120310bed270bf9062","sha256:b52b0a480e223c205f2375b4bb1d7e91df1ea9ce26fce8910e8277fda09f7c32"],"state_sha256":"bd555d785fc593fb44bfc3c71b2665bd1fbc82d30a77d043212cc5c3860b1052"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5/9+wlXEom0+h+W41MgeGHGmhmmXt7uOBzGm8rkdt3M0lLFgwqUvq/4g0KpkI9lrb+j22JZBWvSPcCOurOOmCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T05:55:23.835992Z","bundle_sha256":"8a7b3966a725277e151b50e057a0b7728a0d910ca174a08f5e5e5693db4644ac"}}