{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:SNFW3OFXBCFNV32IWWMBPCFGOS","short_pith_number":"pith:SNFW3OFX","schema_version":"1.0","canonical_sha256":"934b6db8b7088adaef48b5981788a6749321f6504e378603b34f379b4596491c","source":{"kind":"arxiv","id":"2205.11640","version":2},"attestation_state":"computed","paper":{"title":"Generalization Gap in Amortized Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"David Barber, Mingtian Zhang, Peter Hayes","submitted_at":"2022-05-23T21:28:47Z","abstract_excerpt":"The ability of likelihood-based probabilistic models to generalize to unseen data is central to many machine learning applications such as lossless compression. In this work, we study the generalization of a popular class of probabilistic model - the Variational Auto-Encoder (VAE). We discuss the two generalization gaps that affect VAEs and show that overfitting is usually dominated by amortized inference. Based on this observation, we propose a new training objective that improves the generalization of amortized inference. We demonstrate how our method can improve performance in the context o"},"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":"2205.11640","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2022-05-23T21:28:47Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"33e282dbc8c5a14d6d3822338696dda544485c75dfc4582c3b6ec7dbdf157b4e","abstract_canon_sha256":"621da813e119aa5c89386f37e1026103c35bbe7441e8ec1856b2f0cf4f58443c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:06:58.502973Z","signature_b64":"FYdy4HsLcKkQGHFD/wfBaO/bNZI3G9FkAmq3cnZPjk9Rh1YTSz9jPcgyaRuiMvLIGv4aWJwkqK+44QSJepNhCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"934b6db8b7088adaef48b5981788a6749321f6504e378603b34f379b4596491c","last_reissued_at":"2026-07-05T05:06:58.502515Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:06:58.502515Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generalization Gap in Amortized Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"David Barber, Mingtian Zhang, Peter Hayes","submitted_at":"2022-05-23T21:28:47Z","abstract_excerpt":"The ability of likelihood-based probabilistic models to generalize to unseen data is central to many machine learning applications such as lossless compression. In this work, we study the generalization of a popular class of probabilistic model - the Variational Auto-Encoder (VAE). We discuss the two generalization gaps that affect VAEs and show that overfitting is usually dominated by amortized inference. Based on this observation, we propose a new training objective that improves the generalization of amortized inference. We demonstrate how our method can improve performance in the context o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.11640","kind":"arxiv","version":2},"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/2205.11640/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":"2205.11640","created_at":"2026-07-05T05:06:58.502571+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.11640v2","created_at":"2026-07-05T05:06:58.502571+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.11640","created_at":"2026-07-05T05:06:58.502571+00:00"},{"alias_kind":"pith_short_12","alias_value":"SNFW3OFXBCFN","created_at":"2026-07-05T05:06:58.502571+00:00"},{"alias_kind":"pith_short_16","alias_value":"SNFW3OFXBCFNV32I","created_at":"2026-07-05T05:06:58.502571+00:00"},{"alias_kind":"pith_short_8","alias_value":"SNFW3OFX","created_at":"2026-07-05T05:06:58.502571+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SNFW3OFXBCFNV32IWWMBPCFGOS","json":"https://pith.science/pith/SNFW3OFXBCFNV32IWWMBPCFGOS.json","graph_json":"https://pith.science/api/pith-number/SNFW3OFXBCFNV32IWWMBPCFGOS/graph.json","events_json":"https://pith.science/api/pith-number/SNFW3OFXBCFNV32IWWMBPCFGOS/events.json","paper":"https://pith.science/paper/SNFW3OFX"},"agent_actions":{"view_html":"https://pith.science/pith/SNFW3OFXBCFNV32IWWMBPCFGOS","download_json":"https://pith.science/pith/SNFW3OFXBCFNV32IWWMBPCFGOS.json","view_paper":"https://pith.science/paper/SNFW3OFX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.11640&json=true","fetch_graph":"https://pith.science/api/pith-number/SNFW3OFXBCFNV32IWWMBPCFGOS/graph.json","fetch_events":"https://pith.science/api/pith-number/SNFW3OFXBCFNV32IWWMBPCFGOS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SNFW3OFXBCFNV32IWWMBPCFGOS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SNFW3OFXBCFNV32IWWMBPCFGOS/action/storage_attestation","attest_author":"https://pith.science/pith/SNFW3OFXBCFNV32IWWMBPCFGOS/action/author_attestation","sign_citation":"https://pith.science/pith/SNFW3OFXBCFNV32IWWMBPCFGOS/action/citation_signature","submit_replication":"https://pith.science/pith/SNFW3OFXBCFNV32IWWMBPCFGOS/action/replication_record"}},"created_at":"2026-07-05T05:06:58.502571+00:00","updated_at":"2026-07-05T05:06:58.502571+00:00"}