{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:JVBXF26IMGVPAEW7ZNM3DTS2MW","short_pith_number":"pith:JVBXF26I","schema_version":"1.0","canonical_sha256":"4d4372ebc861aaf012dfcb59b1ce5a65bfdc0b39f9280ef462d99399532cf257","source":{"kind":"arxiv","id":"2202.01627","version":3},"attestation_state":"computed","paper":{"title":"Non-Vacuous Generalisation Bounds for Shallow Neural Networks","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Benjamin Guedj, Felix Biggs","submitted_at":"2022-02-03T14:59:51Z","abstract_excerpt":"We focus on a specific class of shallow neural networks with a single hidden layer, namely those with $L_2$-normalised data and either a sigmoid-shaped Gaussian error function (\"erf\") activation or a Gaussian Error Linear Unit (GELU) activation. For these networks, we derive new generalisation bounds through the PAC-Bayesian theory; unlike most existing such bounds they apply to neural networks with deterministic rather than randomised parameters. Our bounds are empirically non-vacuous when the network is trained with vanilla stochastic gradient descent on MNIST and Fashion-MNIST."},"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":"2202.01627","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2022-02-03T14:59:51Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"1f0fd2f8c0a789119e82108ba73404abec5c1c74fccddd22e4ba0b56fcef7a43","abstract_canon_sha256":"1911b7ffdadb7e1de2254382d9195049a3223e792c8413c74849f2deea0817b7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:08:26.289445Z","signature_b64":"IwcwOVLhKopZlg7S7TLAROYNX5D3rQCz+RmbxtEo7l9cTQYjr/vAFlCrVQTVeLyvYgt4xqL3zcC/2f+ZhwtZCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4d4372ebc861aaf012dfcb59b1ce5a65bfdc0b39f9280ef462d99399532cf257","last_reissued_at":"2026-07-05T05:08:26.289073Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:08:26.289073Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Non-Vacuous Generalisation Bounds for Shallow Neural Networks","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Benjamin Guedj, Felix Biggs","submitted_at":"2022-02-03T14:59:51Z","abstract_excerpt":"We focus on a specific class of shallow neural networks with a single hidden layer, namely those with $L_2$-normalised data and either a sigmoid-shaped Gaussian error function (\"erf\") activation or a Gaussian Error Linear Unit (GELU) activation. For these networks, we derive new generalisation bounds through the PAC-Bayesian theory; unlike most existing such bounds they apply to neural networks with deterministic rather than randomised parameters. Our bounds are empirically non-vacuous when the network is trained with vanilla stochastic gradient descent on MNIST and Fashion-MNIST."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.01627","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/2202.01627/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":"2202.01627","created_at":"2026-07-05T05:08:26.289132+00:00"},{"alias_kind":"arxiv_version","alias_value":"2202.01627v3","created_at":"2026-07-05T05:08:26.289132+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.01627","created_at":"2026-07-05T05:08:26.289132+00:00"},{"alias_kind":"pith_short_12","alias_value":"JVBXF26IMGVP","created_at":"2026-07-05T05:08:26.289132+00:00"},{"alias_kind":"pith_short_16","alias_value":"JVBXF26IMGVPAEW7","created_at":"2026-07-05T05:08:26.289132+00:00"},{"alias_kind":"pith_short_8","alias_value":"JVBXF26I","created_at":"2026-07-05T05:08:26.289132+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.14288","citing_title":"On the Sample Complexity of One Hidden Layer Networks with Equivariance, Locality and Weight Sharing","ref_index":30,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JVBXF26IMGVPAEW7ZNM3DTS2MW","json":"https://pith.science/pith/JVBXF26IMGVPAEW7ZNM3DTS2MW.json","graph_json":"https://pith.science/api/pith-number/JVBXF26IMGVPAEW7ZNM3DTS2MW/graph.json","events_json":"https://pith.science/api/pith-number/JVBXF26IMGVPAEW7ZNM3DTS2MW/events.json","paper":"https://pith.science/paper/JVBXF26I"},"agent_actions":{"view_html":"https://pith.science/pith/JVBXF26IMGVPAEW7ZNM3DTS2MW","download_json":"https://pith.science/pith/JVBXF26IMGVPAEW7ZNM3DTS2MW.json","view_paper":"https://pith.science/paper/JVBXF26I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2202.01627&json=true","fetch_graph":"https://pith.science/api/pith-number/JVBXF26IMGVPAEW7ZNM3DTS2MW/graph.json","fetch_events":"https://pith.science/api/pith-number/JVBXF26IMGVPAEW7ZNM3DTS2MW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JVBXF26IMGVPAEW7ZNM3DTS2MW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JVBXF26IMGVPAEW7ZNM3DTS2MW/action/storage_attestation","attest_author":"https://pith.science/pith/JVBXF26IMGVPAEW7ZNM3DTS2MW/action/author_attestation","sign_citation":"https://pith.science/pith/JVBXF26IMGVPAEW7ZNM3DTS2MW/action/citation_signature","submit_replication":"https://pith.science/pith/JVBXF26IMGVPAEW7ZNM3DTS2MW/action/replication_record"}},"created_at":"2026-07-05T05:08:26.289132+00:00","updated_at":"2026-07-05T05:08:26.289132+00:00"}