{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:4JXYY5EY6MPIW7KN77IYRSK5VO","short_pith_number":"pith:4JXYY5EY","schema_version":"1.0","canonical_sha256":"e26f8c7498f31e8b7d4dffd188c95dab88b837d82fea17fb068fe677826d138b","source":{"kind":"arxiv","id":"1911.04594","version":6},"attestation_state":"computed","paper":{"title":"Rate-Regularization and Generalization in VAEs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alican Bozkurt, Babak Esmaeili, Dana H. Brooks, Jan-Willem van de Meent, Jean-Baptiste Tristan, Jennifer G. Dy","submitted_at":"2019-11-11T23:06:40Z","abstract_excerpt":"Variational autoencoders optimize an objective that combines a reconstruction loss (the distortion) and a KL term (the rate). The rate is an upper bound on the mutual information, which is often interpreted as a regularizer that controls the degree of compression. We here examine whether inclusion of the rate also acts as an inductive bias that improves generalization. We perform rate-distortion analyses that control the strength of the rate term, the network capacity, and the difficulty of the generalization problem. Decreasing the strength of the rate paradoxically improves generalization in"},"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":"1911.04594","kind":"arxiv","version":6},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-11-11T23:06:40Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"99d7d0645fd0e3f47e9da03e75c20fdd50aaf1937cf5dc7fe17c732627a0a941","abstract_canon_sha256":"efed647f164875c17e6afb949947bbd40e0b3c517ce1a5db513cc5b0050b5d71"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:26:25.934592Z","signature_b64":"gTpaWZftINrVhi9yVUZN7chFU6pyPgMGQKWGKO2BR67KeuDKaKU7wuV1kzjMPFA8RDfjXX1JxNrrvKY3dNULAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e26f8c7498f31e8b7d4dffd188c95dab88b837d82fea17fb068fe677826d138b","last_reissued_at":"2026-07-05T02:26:25.934135Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:26:25.934135Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rate-Regularization and Generalization in VAEs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alican Bozkurt, Babak Esmaeili, Dana H. Brooks, Jan-Willem van de Meent, Jean-Baptiste Tristan, Jennifer G. Dy","submitted_at":"2019-11-11T23:06:40Z","abstract_excerpt":"Variational autoencoders optimize an objective that combines a reconstruction loss (the distortion) and a KL term (the rate). The rate is an upper bound on the mutual information, which is often interpreted as a regularizer that controls the degree of compression. We here examine whether inclusion of the rate also acts as an inductive bias that improves generalization. We perform rate-distortion analyses that control the strength of the rate term, the network capacity, and the difficulty of the generalization problem. Decreasing the strength of the rate paradoxically improves generalization in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1911.04594","kind":"arxiv","version":6},"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/1911.04594/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":"1911.04594","created_at":"2026-07-05T02:26:25.934184+00:00"},{"alias_kind":"arxiv_version","alias_value":"1911.04594v6","created_at":"2026-07-05T02:26:25.934184+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1911.04594","created_at":"2026-07-05T02:26:25.934184+00:00"},{"alias_kind":"pith_short_12","alias_value":"4JXYY5EY6MPI","created_at":"2026-07-05T02:26:25.934184+00:00"},{"alias_kind":"pith_short_16","alias_value":"4JXYY5EY6MPIW7KN","created_at":"2026-07-05T02:26:25.934184+00:00"},{"alias_kind":"pith_short_8","alias_value":"4JXYY5EY","created_at":"2026-07-05T02:26:25.934184+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.10089","citing_title":"Optimizing Latent Dimension Allocation in Hierarchical VAEs: Balancing Attenuation and Information Retention for OOD Detection","ref_index":20,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4JXYY5EY6MPIW7KN77IYRSK5VO","json":"https://pith.science/pith/4JXYY5EY6MPIW7KN77IYRSK5VO.json","graph_json":"https://pith.science/api/pith-number/4JXYY5EY6MPIW7KN77IYRSK5VO/graph.json","events_json":"https://pith.science/api/pith-number/4JXYY5EY6MPIW7KN77IYRSK5VO/events.json","paper":"https://pith.science/paper/4JXYY5EY"},"agent_actions":{"view_html":"https://pith.science/pith/4JXYY5EY6MPIW7KN77IYRSK5VO","download_json":"https://pith.science/pith/4JXYY5EY6MPIW7KN77IYRSK5VO.json","view_paper":"https://pith.science/paper/4JXYY5EY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1911.04594&json=true","fetch_graph":"https://pith.science/api/pith-number/4JXYY5EY6MPIW7KN77IYRSK5VO/graph.json","fetch_events":"https://pith.science/api/pith-number/4JXYY5EY6MPIW7KN77IYRSK5VO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4JXYY5EY6MPIW7KN77IYRSK5VO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4JXYY5EY6MPIW7KN77IYRSK5VO/action/storage_attestation","attest_author":"https://pith.science/pith/4JXYY5EY6MPIW7KN77IYRSK5VO/action/author_attestation","sign_citation":"https://pith.science/pith/4JXYY5EY6MPIW7KN77IYRSK5VO/action/citation_signature","submit_replication":"https://pith.science/pith/4JXYY5EY6MPIW7KN77IYRSK5VO/action/replication_record"}},"created_at":"2026-07-05T02:26:25.934184+00:00","updated_at":"2026-07-05T02:26:25.934184+00:00"}