{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:AAUQ6DFZOANH2SE7YON7PW2X7T","short_pith_number":"pith:AAUQ6DFZ","schema_version":"1.0","canonical_sha256":"00290f0cb9701a7d489fc39bf7db57fcc718bc87ce677391ae30f519613e38ec","source":{"kind":"arxiv","id":"2110.11886","version":2},"attestation_state":"computed","paper":{"title":"Conditionally Gaussian PAC-Bayes","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Arnaud Doucet, Eugenio Clerico, George Deligiannidis","submitted_at":"2021-10-22T16:12:03Z","abstract_excerpt":"Recent studies have empirically investigated different methods to train stochastic neural networks on a classification task by optimising a PAC-Bayesian bound via stochastic gradient descent. Most of these procedures need to replace the misclassification error with a surrogate loss, leading to a mismatch between the optimisation objective and the actual generalisation bound. The present paper proposes a novel training algorithm that optimises the PAC-Bayesian bound, without relying on any surrogate loss. Empirical results show that this approach outperforms currently available PAC-Bayesian tra"},"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":"2110.11886","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-22T16:12:03Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"dbaba6b6fef006839a92bfd7951de9943f1a1883d9c169095f20f03ffcf7b413","abstract_canon_sha256":"0a63ef6a62fe3ca090656f969e9478a198adf6551ac36111c1046951db721305"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:36:20.953313Z","signature_b64":"s3jw7FOL+vFzl39jTq6YCdMzaHum+oBRfCGdaAYZfKf2qu8PI616+RFw8NrwDghIY7FAaUBSFU/3imaeBrKZCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"00290f0cb9701a7d489fc39bf7db57fcc718bc87ce677391ae30f519613e38ec","last_reissued_at":"2026-07-05T04:36:20.952825Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:36:20.952825Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Conditionally Gaussian PAC-Bayes","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Arnaud Doucet, Eugenio Clerico, George Deligiannidis","submitted_at":"2021-10-22T16:12:03Z","abstract_excerpt":"Recent studies have empirically investigated different methods to train stochastic neural networks on a classification task by optimising a PAC-Bayesian bound via stochastic gradient descent. Most of these procedures need to replace the misclassification error with a surrogate loss, leading to a mismatch between the optimisation objective and the actual generalisation bound. The present paper proposes a novel training algorithm that optimises the PAC-Bayesian bound, without relying on any surrogate loss. Empirical results show that this approach outperforms currently available PAC-Bayesian tra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.11886","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/2110.11886/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":"2110.11886","created_at":"2026-07-05T04:36:20.952883+00:00"},{"alias_kind":"arxiv_version","alias_value":"2110.11886v2","created_at":"2026-07-05T04:36:20.952883+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.11886","created_at":"2026-07-05T04:36:20.952883+00:00"},{"alias_kind":"pith_short_12","alias_value":"AAUQ6DFZOANH","created_at":"2026-07-05T04:36:20.952883+00:00"},{"alias_kind":"pith_short_16","alias_value":"AAUQ6DFZOANH2SE7","created_at":"2026-07-05T04:36:20.952883+00:00"},{"alias_kind":"pith_short_8","alias_value":"AAUQ6DFZ","created_at":"2026-07-05T04:36:20.952883+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/AAUQ6DFZOANH2SE7YON7PW2X7T","json":"https://pith.science/pith/AAUQ6DFZOANH2SE7YON7PW2X7T.json","graph_json":"https://pith.science/api/pith-number/AAUQ6DFZOANH2SE7YON7PW2X7T/graph.json","events_json":"https://pith.science/api/pith-number/AAUQ6DFZOANH2SE7YON7PW2X7T/events.json","paper":"https://pith.science/paper/AAUQ6DFZ"},"agent_actions":{"view_html":"https://pith.science/pith/AAUQ6DFZOANH2SE7YON7PW2X7T","download_json":"https://pith.science/pith/AAUQ6DFZOANH2SE7YON7PW2X7T.json","view_paper":"https://pith.science/paper/AAUQ6DFZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2110.11886&json=true","fetch_graph":"https://pith.science/api/pith-number/AAUQ6DFZOANH2SE7YON7PW2X7T/graph.json","fetch_events":"https://pith.science/api/pith-number/AAUQ6DFZOANH2SE7YON7PW2X7T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AAUQ6DFZOANH2SE7YON7PW2X7T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AAUQ6DFZOANH2SE7YON7PW2X7T/action/storage_attestation","attest_author":"https://pith.science/pith/AAUQ6DFZOANH2SE7YON7PW2X7T/action/author_attestation","sign_citation":"https://pith.science/pith/AAUQ6DFZOANH2SE7YON7PW2X7T/action/citation_signature","submit_replication":"https://pith.science/pith/AAUQ6DFZOANH2SE7YON7PW2X7T/action/replication_record"}},"created_at":"2026-07-05T04:36:20.952883+00:00","updated_at":"2026-07-05T04:36:20.952883+00:00"}