{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:JAWEIQ6A63L2JBEU2HJQXMNGN6","short_pith_number":"pith:JAWEIQ6A","schema_version":"1.0","canonical_sha256":"482c4443c0f6d7a48494d1d30bb1a66faa5a8579aa3d562a2a99665e8bc9cbde","source":{"kind":"arxiv","id":"1901.03416","version":1},"attestation_state":"computed","paper":{"title":"Preventing Posterior Collapse with delta-VAEs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"A\\\"aron van den Oord, Ali Razavi, Ben Poole, Oriol Vinyals","submitted_at":"2019-01-10T22:13:15Z","abstract_excerpt":"Due to the phenomenon of \"posterior collapse,\" current latent variable generative models pose a challenging design choice that either weakens the capacity of the decoder or requires augmenting the objective so it does not only maximize the likelihood of the data. In this paper, we propose an alternative that utilizes the most powerful generative models as decoders, whilst optimising the variational lower bound all while ensuring that the latent variables preserve and encode useful information. Our proposed $\\delta$-VAEs achieve this by constraining the variational family for the posterior to h"},"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":"1901.03416","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-10T22:13:15Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"33a318637143e48e53d614ee187004385d31cf07b4ad70dd6bf271d2d22fd6f0","abstract_canon_sha256":"c4f0802b3c3fb6d587870bc4abc978183a8e52fb7f1b8135de367c729b70505e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:56:31.284438Z","signature_b64":"Y075jKxD7GK/0rRHFGtRn9S6reGb+e5x2zFGpNHg2tC9OlS4OhR+y9ypICiiQWnwOWHS87ZA5qbQVFxZTcAMDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"482c4443c0f6d7a48494d1d30bb1a66faa5a8579aa3d562a2a99665e8bc9cbde","last_reissued_at":"2026-05-17T23:56:31.283960Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:56:31.283960Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Preventing Posterior Collapse with delta-VAEs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"A\\\"aron van den Oord, Ali Razavi, Ben Poole, Oriol Vinyals","submitted_at":"2019-01-10T22:13:15Z","abstract_excerpt":"Due to the phenomenon of \"posterior collapse,\" current latent variable generative models pose a challenging design choice that either weakens the capacity of the decoder or requires augmenting the objective so it does not only maximize the likelihood of the data. In this paper, we propose an alternative that utilizes the most powerful generative models as decoders, whilst optimising the variational lower bound all while ensuring that the latent variables preserve and encode useful information. Our proposed $\\delta$-VAEs achieve this by constraining the variational family for the posterior to h"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1901.03416","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"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":"1901.03416","created_at":"2026-05-17T23:56:31.284034+00:00"},{"alias_kind":"arxiv_version","alias_value":"1901.03416v1","created_at":"2026-05-17T23:56:31.284034+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1901.03416","created_at":"2026-05-17T23:56:31.284034+00:00"},{"alias_kind":"pith_short_12","alias_value":"JAWEIQ6A63L2","created_at":"2026-05-18T12:33:18.533446+00:00"},{"alias_kind":"pith_short_16","alias_value":"JAWEIQ6A63L2JBEU","created_at":"2026-05-18T12:33:18.533446+00:00"},{"alias_kind":"pith_short_8","alias_value":"JAWEIQ6A","created_at":"2026-05-18T12:33:18.533446+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":4,"sample":[{"citing_arxiv_id":"2606.27681","citing_title":"Textual Belief States for World Models: Identifiable Representation Learning Under Strict Mediation","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2606.04935","citing_title":"What Type of Inference is Active Inference?","ref_index":279,"is_internal_anchor":true},{"citing_arxiv_id":"2606.20658","citing_title":"Expected Free Energy-based Planning as Variational Inference","ref_index":256,"is_internal_anchor":true},{"citing_arxiv_id":"1906.09237","citing_title":"Shaping Belief States with Generative Environment Models for RL","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2605.03413","citing_title":"Learning to Theorize the World from Observation","ref_index":185,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JAWEIQ6A63L2JBEU2HJQXMNGN6","json":"https://pith.science/pith/JAWEIQ6A63L2JBEU2HJQXMNGN6.json","graph_json":"https://pith.science/api/pith-number/JAWEIQ6A63L2JBEU2HJQXMNGN6/graph.json","events_json":"https://pith.science/api/pith-number/JAWEIQ6A63L2JBEU2HJQXMNGN6/events.json","paper":"https://pith.science/paper/JAWEIQ6A"},"agent_actions":{"view_html":"https://pith.science/pith/JAWEIQ6A63L2JBEU2HJQXMNGN6","download_json":"https://pith.science/pith/JAWEIQ6A63L2JBEU2HJQXMNGN6.json","view_paper":"https://pith.science/paper/JAWEIQ6A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1901.03416&json=true","fetch_graph":"https://pith.science/api/pith-number/JAWEIQ6A63L2JBEU2HJQXMNGN6/graph.json","fetch_events":"https://pith.science/api/pith-number/JAWEIQ6A63L2JBEU2HJQXMNGN6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JAWEIQ6A63L2JBEU2HJQXMNGN6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JAWEIQ6A63L2JBEU2HJQXMNGN6/action/storage_attestation","attest_author":"https://pith.science/pith/JAWEIQ6A63L2JBEU2HJQXMNGN6/action/author_attestation","sign_citation":"https://pith.science/pith/JAWEIQ6A63L2JBEU2HJQXMNGN6/action/citation_signature","submit_replication":"https://pith.science/pith/JAWEIQ6A63L2JBEU2HJQXMNGN6/action/replication_record"}},"created_at":"2026-05-17T23:56:31.284034+00:00","updated_at":"2026-05-17T23:56:31.284034+00:00"}