{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:CDEWRWDBKXBURSL6BNRGGZRTBQ","short_pith_number":"pith:CDEWRWDB","schema_version":"1.0","canonical_sha256":"10c968d86155c348c97e0b626366330c26347b78b2bfbc2134da98db2fe20e6b","source":{"kind":"arxiv","id":"2402.06578","version":3},"attestation_state":"computed","paper":{"title":"On the Universality of Volume-Preserving and Coupling-Based Normalizing Flows","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Christoph Schn\\\"orr, Felix Draxler, Stefan Wahl, Ullrich K\\\"othe","submitted_at":"2024-02-09T17:51:43Z","abstract_excerpt":"We present a novel theoretical framework for understanding the expressive power of normalizing flows. Despite their prevalence in scientific applications, a comprehensive understanding of flows remains elusive due to their restricted architectures. Existing theorems fall short as they require the use of arbitrarily ill-conditioned neural networks, limiting practical applicability. We propose a distributional universality theorem for well-conditioned coupling-based normalizing flows such as RealNVP. In addition, we show that volume-preserving normalizing flows are not universal, what distributi"},"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":"2402.06578","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-09T17:51:43Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"2bc0ea23c2db975929cade18454931f305f885e85312f02eb3706f2def628c09","abstract_canon_sha256":"1b293b76f34368727d9a1cf23dda2eafa20e2231cc0a8c27a2d215e6605f22cc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:06:39.188713Z","signature_b64":"dbcNVUIxZoZHe4QzNK66vW+F+xrhL5fksZVvlXsqVT7u3Fb9yUFwtFKkYS2mLSG2cx8fS9aGHP5pxn+oAyisAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"10c968d86155c348c97e0b626366330c26347b78b2bfbc2134da98db2fe20e6b","last_reissued_at":"2026-07-05T10:06:39.188274Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:06:39.188274Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Universality of Volume-Preserving and Coupling-Based Normalizing Flows","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Christoph Schn\\\"orr, Felix Draxler, Stefan Wahl, Ullrich K\\\"othe","submitted_at":"2024-02-09T17:51:43Z","abstract_excerpt":"We present a novel theoretical framework for understanding the expressive power of normalizing flows. Despite their prevalence in scientific applications, a comprehensive understanding of flows remains elusive due to their restricted architectures. Existing theorems fall short as they require the use of arbitrarily ill-conditioned neural networks, limiting practical applicability. We propose a distributional universality theorem for well-conditioned coupling-based normalizing flows such as RealNVP. In addition, we show that volume-preserving normalizing flows are not universal, what distributi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.06578","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/2402.06578/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":"2402.06578","created_at":"2026-07-05T10:06:39.188336+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.06578v3","created_at":"2026-07-05T10:06:39.188336+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.06578","created_at":"2026-07-05T10:06:39.188336+00:00"},{"alias_kind":"pith_short_12","alias_value":"CDEWRWDBKXBU","created_at":"2026-07-05T10:06:39.188336+00:00"},{"alias_kind":"pith_short_16","alias_value":"CDEWRWDBKXBURSL6","created_at":"2026-07-05T10:06:39.188336+00:00"},{"alias_kind":"pith_short_8","alias_value":"CDEWRWDB","created_at":"2026-07-05T10:06:39.188336+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26016","citing_title":"MIMFlow: Integrating Masked Image Modeling with Normalizing Flows for End-to-End Image Generation","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21385","citing_title":"Unsupervised Disentanglement Without Compromises : How Functional Orthogonality Enforces Identifiability","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2606.26016","citing_title":"MIMFlow: Integrating Masked Image Modeling with Normalizing Flows for End-to-End Image Generation","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2306.10430","citing_title":"Variational Sequential Optimal Experimental Design using Reinforcement Learning","ref_index":79,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08582","citing_title":"Multivariate Time Series Anomaly Detection via Dual-Branch Reconstruction and Autoregressive Flow-based Residual Density Estimation","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06172","citing_title":"Expressivity of Bi-Lipschitz Normalizing Flows: A Score-Based Diffusion Perspective","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CDEWRWDBKXBURSL6BNRGGZRTBQ","json":"https://pith.science/pith/CDEWRWDBKXBURSL6BNRGGZRTBQ.json","graph_json":"https://pith.science/api/pith-number/CDEWRWDBKXBURSL6BNRGGZRTBQ/graph.json","events_json":"https://pith.science/api/pith-number/CDEWRWDBKXBURSL6BNRGGZRTBQ/events.json","paper":"https://pith.science/paper/CDEWRWDB"},"agent_actions":{"view_html":"https://pith.science/pith/CDEWRWDBKXBURSL6BNRGGZRTBQ","download_json":"https://pith.science/pith/CDEWRWDBKXBURSL6BNRGGZRTBQ.json","view_paper":"https://pith.science/paper/CDEWRWDB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.06578&json=true","fetch_graph":"https://pith.science/api/pith-number/CDEWRWDBKXBURSL6BNRGGZRTBQ/graph.json","fetch_events":"https://pith.science/api/pith-number/CDEWRWDBKXBURSL6BNRGGZRTBQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CDEWRWDBKXBURSL6BNRGGZRTBQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CDEWRWDBKXBURSL6BNRGGZRTBQ/action/storage_attestation","attest_author":"https://pith.science/pith/CDEWRWDBKXBURSL6BNRGGZRTBQ/action/author_attestation","sign_citation":"https://pith.science/pith/CDEWRWDBKXBURSL6BNRGGZRTBQ/action/citation_signature","submit_replication":"https://pith.science/pith/CDEWRWDBKXBURSL6BNRGGZRTBQ/action/replication_record"}},"created_at":"2026-07-05T10:06:39.188336+00:00","updated_at":"2026-07-05T10:06:39.188336+00:00"}