{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:MWRKGHNN6HXYJNN4JEFI3NWJFD","short_pith_number":"pith:MWRKGHNN","schema_version":"1.0","canonical_sha256":"65a2a31dadf1ef84b5bc490a8db6c928c986926fd6c7c62fe8c172f83ca9b616","source":{"kind":"arxiv","id":"2307.11018","version":4},"attestation_state":"computed","paper":{"title":"Amortized Variational Inference: When and Why?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Charles C. Margossian, David M. Blei","submitted_at":"2023-07-20T16:45:22Z","abstract_excerpt":"In a probabilistic latent variable model, factorized (or mean-field) variational inference (F-VI) fits a separate parametric distribution for each latent variable. Amortized variational inference (A-VI) instead learns a common inference function, which maps each observation to its corresponding latent variable's approximate posterior. Typically, A-VI is used as a step in the training of variational autoencoders, however it stands to reason that A-VI could also be used as a general alternative to F-VI. In this paper we study when and why A-VI can be used for approximate Bayesian inference. We d"},"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":"2307.11018","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2023-07-20T16:45:22Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"109cdf041305daf588d4168163c8f7c2f8928033d161d05e1fdb157093356a93","abstract_canon_sha256":"afc67d8448f892f5f3fb14eb6d5a72925f19f017a90f8b5b59fbc9d1f35476b5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:22:32.529365Z","signature_b64":"7a8w7PNaJiBPFCa3OkiaN8M74lKmdXWbDlLXDfg9DzhUBXNAsiIdfgjrsN0/c0J+JygtBy0SD4D+osBZ7t1ZAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"65a2a31dadf1ef84b5bc490a8db6c928c986926fd6c7c62fe8c172f83ca9b616","last_reissued_at":"2026-07-05T08:22:32.528745Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:22:32.528745Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Amortized Variational Inference: When and Why?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Charles C. Margossian, David M. Blei","submitted_at":"2023-07-20T16:45:22Z","abstract_excerpt":"In a probabilistic latent variable model, factorized (or mean-field) variational inference (F-VI) fits a separate parametric distribution for each latent variable. Amortized variational inference (A-VI) instead learns a common inference function, which maps each observation to its corresponding latent variable's approximate posterior. Typically, A-VI is used as a step in the training of variational autoencoders, however it stands to reason that A-VI could also be used as a general alternative to F-VI. In this paper we study when and why A-VI can be used for approximate Bayesian inference. We d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2307.11018","kind":"arxiv","version":4},"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/2307.11018/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":"2307.11018","created_at":"2026-07-05T08:22:32.528809+00:00"},{"alias_kind":"arxiv_version","alias_value":"2307.11018v4","created_at":"2026-07-05T08:22:32.528809+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2307.11018","created_at":"2026-07-05T08:22:32.528809+00:00"},{"alias_kind":"pith_short_12","alias_value":"MWRKGHNN6HXY","created_at":"2026-07-05T08:22:32.528809+00:00"},{"alias_kind":"pith_short_16","alias_value":"MWRKGHNN6HXYJNN4","created_at":"2026-07-05T08:22:32.528809+00:00"},{"alias_kind":"pith_short_8","alias_value":"MWRKGHNN","created_at":"2026-07-05T08:22:32.528809+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.19867","citing_title":"Deep Active Inference Agents for Delayed and Long-Horizon Environments","ref_index":48,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MWRKGHNN6HXYJNN4JEFI3NWJFD","json":"https://pith.science/pith/MWRKGHNN6HXYJNN4JEFI3NWJFD.json","graph_json":"https://pith.science/api/pith-number/MWRKGHNN6HXYJNN4JEFI3NWJFD/graph.json","events_json":"https://pith.science/api/pith-number/MWRKGHNN6HXYJNN4JEFI3NWJFD/events.json","paper":"https://pith.science/paper/MWRKGHNN"},"agent_actions":{"view_html":"https://pith.science/pith/MWRKGHNN6HXYJNN4JEFI3NWJFD","download_json":"https://pith.science/pith/MWRKGHNN6HXYJNN4JEFI3NWJFD.json","view_paper":"https://pith.science/paper/MWRKGHNN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2307.11018&json=true","fetch_graph":"https://pith.science/api/pith-number/MWRKGHNN6HXYJNN4JEFI3NWJFD/graph.json","fetch_events":"https://pith.science/api/pith-number/MWRKGHNN6HXYJNN4JEFI3NWJFD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MWRKGHNN6HXYJNN4JEFI3NWJFD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MWRKGHNN6HXYJNN4JEFI3NWJFD/action/storage_attestation","attest_author":"https://pith.science/pith/MWRKGHNN6HXYJNN4JEFI3NWJFD/action/author_attestation","sign_citation":"https://pith.science/pith/MWRKGHNN6HXYJNN4JEFI3NWJFD/action/citation_signature","submit_replication":"https://pith.science/pith/MWRKGHNN6HXYJNN4JEFI3NWJFD/action/replication_record"}},"created_at":"2026-07-05T08:22:32.528809+00:00","updated_at":"2026-07-05T08:22:32.528809+00:00"}