{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:6SCB5XOGIDQMMZ4VXMVLGC3K42","short_pith_number":"pith:6SCB5XOG","schema_version":"1.0","canonical_sha256":"f4841eddc640e0c66795bb2ab30b6ae6acc66f01568f40ce50890bf2e394dc9e","source":{"kind":"arxiv","id":"2112.10852","version":3},"attestation_state":"computed","paper":{"title":"The effective noise of Stochastic Gradient Descent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cond-mat.dis-nn","authors_text":"Francesca Mignacco, Pierfrancesco Urbani","submitted_at":"2021-12-20T20:46:19Z","abstract_excerpt":"Stochastic Gradient Descent (SGD) is the workhorse algorithm of deep learning technology. At each step of the training phase, a mini batch of samples is drawn from the training dataset and the weights of the neural network are adjusted according to the performance on this specific subset of examples. The mini-batch sampling procedure introduces a stochastic dynamics to the gradient descent, with a non-trivial state-dependent noise. We characterize the stochasticity of SGD and a recently-introduced variant, \\emph{persistent} SGD, in a prototypical neural network model. In the under-parametrized"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2112.10852","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cond-mat.dis-nn","submitted_at":"2021-12-20T20:46:19Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"c74f867a30c556aec174cfea7bcd1943a47b212a13534b8148dae1459c2cf16f","abstract_canon_sha256":"8aa056ca6137e8b184e69042b7e2bd9fb11a1875d841ad93d4e260ae708d3706"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:54:49.257396Z","signature_b64":"gQtWmlje2y5QBUfojTPUmU/NscOfrgcHzXstF6I0vc/K21UCE6hwVECDbkN9IUqh2ub6tNNiT81J+SeRmMsFAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f4841eddc640e0c66795bb2ab30b6ae6acc66f01568f40ce50890bf2e394dc9e","last_reissued_at":"2026-07-05T04:54:49.256924Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:54:49.256924Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The effective noise of Stochastic Gradient Descent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cond-mat.dis-nn","authors_text":"Francesca Mignacco, Pierfrancesco Urbani","submitted_at":"2021-12-20T20:46:19Z","abstract_excerpt":"Stochastic Gradient Descent (SGD) is the workhorse algorithm of deep learning technology. At each step of the training phase, a mini batch of samples is drawn from the training dataset and the weights of the neural network are adjusted according to the performance on this specific subset of examples. The mini-batch sampling procedure introduces a stochastic dynamics to the gradient descent, with a non-trivial state-dependent noise. We characterize the stochasticity of SGD and a recently-introduced variant, \\emph{persistent} SGD, in a prototypical neural network model. In the under-parametrized"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2112.10852","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/2112.10852/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2112.10852","created_at":"2026-07-05T04:54:49.256984+00:00"},{"alias_kind":"arxiv_version","alias_value":"2112.10852v3","created_at":"2026-07-05T04:54:49.256984+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2112.10852","created_at":"2026-07-05T04:54:49.256984+00:00"},{"alias_kind":"pith_short_12","alias_value":"6SCB5XOGIDQM","created_at":"2026-07-05T04:54:49.256984+00:00"},{"alias_kind":"pith_short_16","alias_value":"6SCB5XOGIDQMMZ4V","created_at":"2026-07-05T04:54:49.256984+00:00"},{"alias_kind":"pith_short_8","alias_value":"6SCB5XOG","created_at":"2026-07-05T04:54:49.256984+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6SCB5XOGIDQMMZ4VXMVLGC3K42","json":"https://pith.science/pith/6SCB5XOGIDQMMZ4VXMVLGC3K42.json","graph_json":"https://pith.science/api/pith-number/6SCB5XOGIDQMMZ4VXMVLGC3K42/graph.json","events_json":"https://pith.science/api/pith-number/6SCB5XOGIDQMMZ4VXMVLGC3K42/events.json","paper":"https://pith.science/paper/6SCB5XOG"},"agent_actions":{"view_html":"https://pith.science/pith/6SCB5XOGIDQMMZ4VXMVLGC3K42","download_json":"https://pith.science/pith/6SCB5XOGIDQMMZ4VXMVLGC3K42.json","view_paper":"https://pith.science/paper/6SCB5XOG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2112.10852&json=true","fetch_graph":"https://pith.science/api/pith-number/6SCB5XOGIDQMMZ4VXMVLGC3K42/graph.json","fetch_events":"https://pith.science/api/pith-number/6SCB5XOGIDQMMZ4VXMVLGC3K42/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6SCB5XOGIDQMMZ4VXMVLGC3K42/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6SCB5XOGIDQMMZ4VXMVLGC3K42/action/storage_attestation","attest_author":"https://pith.science/pith/6SCB5XOGIDQMMZ4VXMVLGC3K42/action/author_attestation","sign_citation":"https://pith.science/pith/6SCB5XOGIDQMMZ4VXMVLGC3K42/action/citation_signature","submit_replication":"https://pith.science/pith/6SCB5XOGIDQMMZ4VXMVLGC3K42/action/replication_record"}},"created_at":"2026-07-05T04:54:49.256984+00:00","updated_at":"2026-07-05T04:54:49.256984+00:00"}