{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:KG6OFTNKJPCZILOF36DV5GQBPE","short_pith_number":"pith:KG6OFTNK","schema_version":"1.0","canonical_sha256":"51bce2cdaa4bc5942dc5df875e9a017915c641bf3508118c93fe286588638766","source":{"kind":"arxiv","id":"1909.03765","version":1},"attestation_state":"computed","paper":{"title":"Balancing Reconstruction Quality and Regularisation in ELBO for VAEs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Niki Trigoni, Ronald Clark, Shuyu Lin, Stephen Roberts","submitted_at":"2019-09-09T11:18:52Z","abstract_excerpt":"A trade-off exists between reconstruction quality and the prior regularisation in the Evidence Lower Bound (ELBO) loss that Variational Autoencoder (VAE) models use for learning. There are few satisfactory approaches to deal with a balance between the prior and reconstruction objective, with most methods dealing with this problem through heuristics. In this paper, we show that the noise variance (often set as a fixed value) in the Gaussian likelihood p(x|z) for real-valued data can naturally act to provide such a balance. By learning this noise variance so as to maximise the ELBO loss, we auto"},"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":"1909.03765","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-09-09T11:18:52Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"8d50abc66fe91b606e669aa343f3055b39895de6ef2e0068f416b44d3d71a450","abstract_canon_sha256":"d08422ce9d4bd4db0e3c80604dc7a76894b1c424425ae01461048732e6ad1724"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:03:08.504050Z","signature_b64":"UqKOdCXFy9Fwo1smY5s6BbDMLOt8vXfuDcnaPtO0KpqzCy7J1pl9qBIQb2dH+tS1xQmgZhroGKwHjDedi8MRAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"51bce2cdaa4bc5942dc5df875e9a017915c641bf3508118c93fe286588638766","last_reissued_at":"2026-07-05T00:03:08.503622Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:03:08.503622Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Balancing Reconstruction Quality and Regularisation in ELBO for VAEs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Niki Trigoni, Ronald Clark, Shuyu Lin, Stephen Roberts","submitted_at":"2019-09-09T11:18:52Z","abstract_excerpt":"A trade-off exists between reconstruction quality and the prior regularisation in the Evidence Lower Bound (ELBO) loss that Variational Autoencoder (VAE) models use for learning. There are few satisfactory approaches to deal with a balance between the prior and reconstruction objective, with most methods dealing with this problem through heuristics. In this paper, we show that the noise variance (often set as a fixed value) in the Gaussian likelihood p(x|z) for real-valued data can naturally act to provide such a balance. By learning this noise variance so as to maximise the ELBO loss, we auto"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1909.03765","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/1909.03765/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":"1909.03765","created_at":"2026-07-05T00:03:08.503679+00:00"},{"alias_kind":"arxiv_version","alias_value":"1909.03765v1","created_at":"2026-07-05T00:03:08.503679+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1909.03765","created_at":"2026-07-05T00:03:08.503679+00:00"},{"alias_kind":"pith_short_12","alias_value":"KG6OFTNKJPCZ","created_at":"2026-07-05T00:03:08.503679+00:00"},{"alias_kind":"pith_short_16","alias_value":"KG6OFTNKJPCZILOF","created_at":"2026-07-05T00:03:08.503679+00:00"},{"alias_kind":"pith_short_8","alias_value":"KG6OFTNK","created_at":"2026-07-05T00:03:08.503679+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26520","citing_title":"Multipath Adaptive Gated Bottleneck Latent ODE with Raman Data Fusion for Cell Culture Process Forecasting","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2507.01976","citing_title":"A Comprehensive Survey on Network Traffic Synthesis: From Statistical Models to Deep Learning","ref_index":86,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KG6OFTNKJPCZILOF36DV5GQBPE","json":"https://pith.science/pith/KG6OFTNKJPCZILOF36DV5GQBPE.json","graph_json":"https://pith.science/api/pith-number/KG6OFTNKJPCZILOF36DV5GQBPE/graph.json","events_json":"https://pith.science/api/pith-number/KG6OFTNKJPCZILOF36DV5GQBPE/events.json","paper":"https://pith.science/paper/KG6OFTNK"},"agent_actions":{"view_html":"https://pith.science/pith/KG6OFTNKJPCZILOF36DV5GQBPE","download_json":"https://pith.science/pith/KG6OFTNKJPCZILOF36DV5GQBPE.json","view_paper":"https://pith.science/paper/KG6OFTNK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1909.03765&json=true","fetch_graph":"https://pith.science/api/pith-number/KG6OFTNKJPCZILOF36DV5GQBPE/graph.json","fetch_events":"https://pith.science/api/pith-number/KG6OFTNKJPCZILOF36DV5GQBPE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KG6OFTNKJPCZILOF36DV5GQBPE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KG6OFTNKJPCZILOF36DV5GQBPE/action/storage_attestation","attest_author":"https://pith.science/pith/KG6OFTNKJPCZILOF36DV5GQBPE/action/author_attestation","sign_citation":"https://pith.science/pith/KG6OFTNKJPCZILOF36DV5GQBPE/action/citation_signature","submit_replication":"https://pith.science/pith/KG6OFTNKJPCZILOF36DV5GQBPE/action/replication_record"}},"created_at":"2026-07-05T00:03:08.503679+00:00","updated_at":"2026-07-05T00:03:08.503679+00:00"}