{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:YGMVSLAQQWJV4DX7QC2VDERQ6S","short_pith_number":"pith:YGMVSLAQ","schema_version":"1.0","canonical_sha256":"c199592c1085935e0eff80b5519230f48d909752c1bffc8a23942df8471802e8","source":{"kind":"arxiv","id":"1711.00464","version":3},"attestation_state":"computed","paper":{"title":"Fixing a Broken ELBO","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alexander A. Alemi, Ben Poole, Ian Fischer, Joshua V. Dillon, Kevin Murphy, Rif A. Saurous","submitted_at":"2017-11-01T17:58:43Z","abstract_excerpt":"Recent work in unsupervised representation learning has focused on learning deep directed latent-variable models. Fitting these models by maximizing the marginal likelihood or evidence is typically intractable, thus a common approximation is to maximize the evidence lower bound (ELBO) instead. However, maximum likelihood training (whether exact or approximate) does not necessarily result in a good latent representation, as we demonstrate both theoretically and empirically. In particular, we derive variational lower and upper bounds on the mutual information between the input and the latent var"},"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":"1711.00464","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-11-01T17:58:43Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"42690d9522b06889d9c0181a46fe6443b9eebaa444ad2b14a5cfb2bde89d478c","abstract_canon_sha256":"1d879062ad3b9fc95a1bda5da9b2461fc3c8890c71c0e49dfdb4a882e20e23cc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:23:22.893676Z","signature_b64":"bflTPEWp4pgiiW7MCQNKncdBQ4o40ufRmo2ddizpTvkT4IJCWwcBt6L5ZWhEZejbdoO19YyL4j7gtTy+NeFODg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c199592c1085935e0eff80b5519230f48d909752c1bffc8a23942df8471802e8","last_reissued_at":"2026-05-18T00:23:22.893033Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:23:22.893033Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fixing a Broken ELBO","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alexander A. Alemi, Ben Poole, Ian Fischer, Joshua V. Dillon, Kevin Murphy, Rif A. Saurous","submitted_at":"2017-11-01T17:58:43Z","abstract_excerpt":"Recent work in unsupervised representation learning has focused on learning deep directed latent-variable models. Fitting these models by maximizing the marginal likelihood or evidence is typically intractable, thus a common approximation is to maximize the evidence lower bound (ELBO) instead. However, maximum likelihood training (whether exact or approximate) does not necessarily result in a good latent representation, as we demonstrate both theoretically and empirically. In particular, we derive variational lower and upper bounds on the mutual information between the input and the latent var"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1711.00464","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":""},"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":"1711.00464","created_at":"2026-05-18T00:23:22.893148+00:00"},{"alias_kind":"arxiv_version","alias_value":"1711.00464v3","created_at":"2026-05-18T00:23:22.893148+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1711.00464","created_at":"2026-05-18T00:23:22.893148+00:00"},{"alias_kind":"pith_short_12","alias_value":"YGMVSLAQQWJV","created_at":"2026-05-18T12:31:56.362134+00:00"},{"alias_kind":"pith_short_16","alias_value":"YGMVSLAQQWJV4DX7","created_at":"2026-05-18T12:31:56.362134+00:00"},{"alias_kind":"pith_short_8","alias_value":"YGMVSLAQ","created_at":"2026-05-18T12:31:56.362134+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":4,"sample":[{"citing_arxiv_id":"1906.09237","citing_title":"Shaping Belief States with Generative Environment Models for RL","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"1906.09155","citing_title":"Query-based Deep Improvisation","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22691","citing_title":"Posterior Collapse as Automatic Spectral Pruning","ref_index":10,"is_internal_anchor":true},{"citing_arxiv_id":"2605.17085","citing_title":"Taming Audio VAEs via Target-KL Regularization","ref_index":25,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YGMVSLAQQWJV4DX7QC2VDERQ6S","json":"https://pith.science/pith/YGMVSLAQQWJV4DX7QC2VDERQ6S.json","graph_json":"https://pith.science/api/pith-number/YGMVSLAQQWJV4DX7QC2VDERQ6S/graph.json","events_json":"https://pith.science/api/pith-number/YGMVSLAQQWJV4DX7QC2VDERQ6S/events.json","paper":"https://pith.science/paper/YGMVSLAQ"},"agent_actions":{"view_html":"https://pith.science/pith/YGMVSLAQQWJV4DX7QC2VDERQ6S","download_json":"https://pith.science/pith/YGMVSLAQQWJV4DX7QC2VDERQ6S.json","view_paper":"https://pith.science/paper/YGMVSLAQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1711.00464&json=true","fetch_graph":"https://pith.science/api/pith-number/YGMVSLAQQWJV4DX7QC2VDERQ6S/graph.json","fetch_events":"https://pith.science/api/pith-number/YGMVSLAQQWJV4DX7QC2VDERQ6S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YGMVSLAQQWJV4DX7QC2VDERQ6S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YGMVSLAQQWJV4DX7QC2VDERQ6S/action/storage_attestation","attest_author":"https://pith.science/pith/YGMVSLAQQWJV4DX7QC2VDERQ6S/action/author_attestation","sign_citation":"https://pith.science/pith/YGMVSLAQQWJV4DX7QC2VDERQ6S/action/citation_signature","submit_replication":"https://pith.science/pith/YGMVSLAQQWJV4DX7QC2VDERQ6S/action/replication_record"}},"created_at":"2026-05-18T00:23:22.893148+00:00","updated_at":"2026-05-18T00:23:22.893148+00:00"}