{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:L6RR5D5OC7RM4BYPXXVNIA3TTP","short_pith_number":"pith:L6RR5D5O","schema_version":"1.0","canonical_sha256":"5fa31e8fae17e2ce070fbdead403739bcb23e2472ea67bbb200d348ffa0dfd37","source":{"kind":"arxiv","id":"2406.04317","version":3},"attestation_state":"computed","paper":{"title":"Regularized KL-Divergence for Well-Defined Function-Space Variational Inference in Bayesian neural networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Robert Bamler, Tristan Cinquin","submitted_at":"2024-06-06T17:57:49Z","abstract_excerpt":"Bayesian neural networks (BNN) promise to combine the predictive performance of neural networks with principled uncertainty modeling important for safety-critical systems and decision making. However, posterior uncertainty estimates depend on the choice of prior, and finding informative priors in weight-space has proven difficult. This has motivated variational inference (VI) methods that pose priors directly on the function generated by the BNN rather than on weights. In this paper, we address a fundamental issue with such function-space VI approaches pointed out by Burt et al. (2020), who sh"},"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":"2406.04317","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-06T17:57:49Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"a0d1c6155278d6b628f7d27c5ebfad5b16509a18e5f8f974edf1f076f3df8650","abstract_canon_sha256":"7f9769157bb5b17d6abee8d04a47e8365d39dc4bc9d59a36008a34643e10740f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:16:02.421937Z","signature_b64":"iXkfKdK7+0fP9AR1JXEm2Uj9l6w4Sj8BR2KsH/6gCH8IssN9o/S+nfz5QutvaiA18sPlWdaH2ks1r6Q7CupeAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5fa31e8fae17e2ce070fbdead403739bcb23e2472ea67bbb200d348ffa0dfd37","last_reissued_at":"2026-07-05T11:16:02.421451Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:16:02.421451Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Regularized KL-Divergence for Well-Defined Function-Space Variational Inference in Bayesian neural networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Robert Bamler, Tristan Cinquin","submitted_at":"2024-06-06T17:57:49Z","abstract_excerpt":"Bayesian neural networks (BNN) promise to combine the predictive performance of neural networks with principled uncertainty modeling important for safety-critical systems and decision making. However, posterior uncertainty estimates depend on the choice of prior, and finding informative priors in weight-space has proven difficult. This has motivated variational inference (VI) methods that pose priors directly on the function generated by the BNN rather than on weights. In this paper, we address a fundamental issue with such function-space VI approaches pointed out by Burt et al. (2020), who sh"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.04317","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/2406.04317/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":"2406.04317","created_at":"2026-07-05T11:16:02.421513+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.04317v3","created_at":"2026-07-05T11:16:02.421513+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.04317","created_at":"2026-07-05T11:16:02.421513+00:00"},{"alias_kind":"pith_short_12","alias_value":"L6RR5D5OC7RM","created_at":"2026-07-05T11:16:02.421513+00:00"},{"alias_kind":"pith_short_16","alias_value":"L6RR5D5OC7RM4BYP","created_at":"2026-07-05T11:16:02.421513+00:00"},{"alias_kind":"pith_short_8","alias_value":"L6RR5D5O","created_at":"2026-07-05T11:16:02.421513+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25745","citing_title":"Gaussian Mean Field Variational Inference can Overestimate Predictive Variance","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/L6RR5D5OC7RM4BYPXXVNIA3TTP","json":"https://pith.science/pith/L6RR5D5OC7RM4BYPXXVNIA3TTP.json","graph_json":"https://pith.science/api/pith-number/L6RR5D5OC7RM4BYPXXVNIA3TTP/graph.json","events_json":"https://pith.science/api/pith-number/L6RR5D5OC7RM4BYPXXVNIA3TTP/events.json","paper":"https://pith.science/paper/L6RR5D5O"},"agent_actions":{"view_html":"https://pith.science/pith/L6RR5D5OC7RM4BYPXXVNIA3TTP","download_json":"https://pith.science/pith/L6RR5D5OC7RM4BYPXXVNIA3TTP.json","view_paper":"https://pith.science/paper/L6RR5D5O","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.04317&json=true","fetch_graph":"https://pith.science/api/pith-number/L6RR5D5OC7RM4BYPXXVNIA3TTP/graph.json","fetch_events":"https://pith.science/api/pith-number/L6RR5D5OC7RM4BYPXXVNIA3TTP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L6RR5D5OC7RM4BYPXXVNIA3TTP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L6RR5D5OC7RM4BYPXXVNIA3TTP/action/storage_attestation","attest_author":"https://pith.science/pith/L6RR5D5OC7RM4BYPXXVNIA3TTP/action/author_attestation","sign_citation":"https://pith.science/pith/L6RR5D5OC7RM4BYPXXVNIA3TTP/action/citation_signature","submit_replication":"https://pith.science/pith/L6RR5D5OC7RM4BYPXXVNIA3TTP/action/replication_record"}},"created_at":"2026-07-05T11:16:02.421513+00:00","updated_at":"2026-07-05T11:16:02.421513+00:00"}