{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:6KIPGKUJ7MA5G7LQABSAPWBNQA","short_pith_number":"pith:6KIPGKUJ","schema_version":"1.0","canonical_sha256":"f290f32a89fb01d37d70006407d82d800f912d16ed7fb90b7d63f524d2e0c97e","source":{"kind":"arxiv","id":"1804.05862","version":3},"attestation_state":"computed","paper":{"title":"Non-Vacuous Generalization Bounds at the ImageNet Scale: A PAC-Bayesian Compression Approach","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Morgane Austern, Peter Orbanz, Ryan P. Adams, Victor Veitch, Wenda Zhou","submitted_at":"2018-04-16T18:01:12Z","abstract_excerpt":"Modern neural networks are highly overparameterized, with capacity to substantially overfit to training data. Nevertheless, these networks often generalize well in practice. It has also been observed that trained networks can often be \"compressed\" to much smaller representations. The purpose of this paper is to connect these two empirical observations. Our main technical result is a generalization bound for compressed networks based on the compressed size. Combined with off-the-shelf compression algorithms, the bound leads to state of the art generalization guarantees; in particular, we provid"},"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":"1804.05862","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-04-16T18:01:12Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"1d74a8c3a79bdc843ebec98f1fce0e0e858628f460c9a3b7a2950bca26bec03b","abstract_canon_sha256":"044e6dd791d0bee050cc9dc6eb6d916bc8391bc12852db9820734f71444006b3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:52:53.164503Z","signature_b64":"FY0Ka5btJvi/kj/qQcOHHPhXWWmVI/8FZZDZeFuVazjp4WNJGXLZIxhmS9NIvBqQ2k2eouTPzt5uc6dTVxLVDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f290f32a89fb01d37d70006407d82d800f912d16ed7fb90b7d63f524d2e0c97e","last_reissued_at":"2026-05-17T23:52:53.163733Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:52:53.163733Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Non-Vacuous Generalization Bounds at the ImageNet Scale: A PAC-Bayesian Compression Approach","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Morgane Austern, Peter Orbanz, Ryan P. Adams, Victor Veitch, Wenda Zhou","submitted_at":"2018-04-16T18:01:12Z","abstract_excerpt":"Modern neural networks are highly overparameterized, with capacity to substantially overfit to training data. Nevertheless, these networks often generalize well in practice. It has also been observed that trained networks can often be \"compressed\" to much smaller representations. The purpose of this paper is to connect these two empirical observations. Our main technical result is a generalization bound for compressed networks based on the compressed size. Combined with off-the-shelf compression algorithms, the bound leads to state of the art generalization guarantees; in particular, we provid"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1804.05862","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":""},"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":"1804.05862","created_at":"2026-05-17T23:52:53.163860+00:00"},{"alias_kind":"arxiv_version","alias_value":"1804.05862v3","created_at":"2026-05-17T23:52:53.163860+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1804.05862","created_at":"2026-05-17T23:52:53.163860+00:00"},{"alias_kind":"pith_short_12","alias_value":"6KIPGKUJ7MA5","created_at":"2026-05-18T12:32:08.215937+00:00"},{"alias_kind":"pith_short_16","alias_value":"6KIPGKUJ7MA5G7LQ","created_at":"2026-05-18T12:32:08.215937+00:00"},{"alias_kind":"pith_short_8","alias_value":"6KIPGKUJ","created_at":"2026-05-18T12:32:08.215937+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"1906.11148","citing_title":"Chaining Meets Chain Rule: Multilevel Entropic Regularization and Training of Neural Nets","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2604.10553","citing_title":"Topology-Aware PAC-Bayesian Generalization Analysis for Graph Neural Networks","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6KIPGKUJ7MA5G7LQABSAPWBNQA","json":"https://pith.science/pith/6KIPGKUJ7MA5G7LQABSAPWBNQA.json","graph_json":"https://pith.science/api/pith-number/6KIPGKUJ7MA5G7LQABSAPWBNQA/graph.json","events_json":"https://pith.science/api/pith-number/6KIPGKUJ7MA5G7LQABSAPWBNQA/events.json","paper":"https://pith.science/paper/6KIPGKUJ"},"agent_actions":{"view_html":"https://pith.science/pith/6KIPGKUJ7MA5G7LQABSAPWBNQA","download_json":"https://pith.science/pith/6KIPGKUJ7MA5G7LQABSAPWBNQA.json","view_paper":"https://pith.science/paper/6KIPGKUJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1804.05862&json=true","fetch_graph":"https://pith.science/api/pith-number/6KIPGKUJ7MA5G7LQABSAPWBNQA/graph.json","fetch_events":"https://pith.science/api/pith-number/6KIPGKUJ7MA5G7LQABSAPWBNQA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6KIPGKUJ7MA5G7LQABSAPWBNQA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6KIPGKUJ7MA5G7LQABSAPWBNQA/action/storage_attestation","attest_author":"https://pith.science/pith/6KIPGKUJ7MA5G7LQABSAPWBNQA/action/author_attestation","sign_citation":"https://pith.science/pith/6KIPGKUJ7MA5G7LQABSAPWBNQA/action/citation_signature","submit_replication":"https://pith.science/pith/6KIPGKUJ7MA5G7LQABSAPWBNQA/action/replication_record"}},"created_at":"2026-05-17T23:52:53.163860+00:00","updated_at":"2026-05-17T23:52:53.163860+00:00"}