{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:JA6RQZYS2L2ZJ3KDIPY4Q55RMQ","short_pith_number":"pith:JA6RQZYS","schema_version":"1.0","canonical_sha256":"483d186712d2f594ed4343f1c877b1642f1ca7ef270948165f0a75d52421c395","source":{"kind":"arxiv","id":"2312.01133","version":2},"attestation_state":"computed","paper":{"title":"$t^3$-Variational Autoencoder: Learning Heavy-tailed Data with Student's t and Power Divergence","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"HyunJong Lee, Jaehyuk Kwon, Joong-Ho Won, Juno Kim, Mincheol Cho","submitted_at":"2023-12-02T13:14:28Z","abstract_excerpt":"The variational autoencoder (VAE) typically employs a standard normal prior as a regularizer for the probabilistic latent encoder. However, the Gaussian tail often decays too quickly to effectively accommodate the encoded points, failing to preserve crucial structures hidden in the data. In this paper, we explore the use of heavy-tailed models to combat over-regularization. Drawing upon insights from information geometry, we propose $t^3$VAE, a modified VAE framework that incorporates Student's t-distributions for the prior, encoder, and decoder. This results in a joint model distribution of a"},"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":"2312.01133","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2023-12-02T13:14:28Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"2207d8bf97679abd7b4da316e2b506f1eb7daaec0eaaa2173c2d748b4f15b3d5","abstract_canon_sha256":"20aa1a1a25c4a7525a53d8595c7d2df65e6bcf99391e2a424011d36f8286729c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:51:36.015012Z","signature_b64":"sufp9Jg760P+zXZTBYC/6Tp+eRmgVLYJWDaHqhMrXDT55pHMUjQV1TTXNqOWElkLcIstM0IpazBBOL9XGattBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"483d186712d2f594ed4343f1c877b1642f1ca7ef270948165f0a75d52421c395","last_reissued_at":"2026-07-05T07:51:36.014573Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:51:36.014573Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"$t^3$-Variational Autoencoder: Learning Heavy-tailed Data with Student's t and Power Divergence","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"HyunJong Lee, Jaehyuk Kwon, Joong-Ho Won, Juno Kim, Mincheol Cho","submitted_at":"2023-12-02T13:14:28Z","abstract_excerpt":"The variational autoencoder (VAE) typically employs a standard normal prior as a regularizer for the probabilistic latent encoder. However, the Gaussian tail often decays too quickly to effectively accommodate the encoded points, failing to preserve crucial structures hidden in the data. In this paper, we explore the use of heavy-tailed models to combat over-regularization. Drawing upon insights from information geometry, we propose $t^3$VAE, a modified VAE framework that incorporates Student's t-distributions for the prior, encoder, and decoder. This results in a joint model distribution of a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.01133","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/2312.01133/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":"2312.01133","created_at":"2026-07-05T07:51:36.014623+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.01133v2","created_at":"2026-07-05T07:51:36.014623+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.01133","created_at":"2026-07-05T07:51:36.014623+00:00"},{"alias_kind":"pith_short_12","alias_value":"JA6RQZYS2L2Z","created_at":"2026-07-05T07:51:36.014623+00:00"},{"alias_kind":"pith_short_16","alias_value":"JA6RQZYS2L2ZJ3KD","created_at":"2026-07-05T07:51:36.014623+00:00"},{"alias_kind":"pith_short_8","alias_value":"JA6RQZYS","created_at":"2026-07-05T07:51:36.014623+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.01645","citing_title":"Self-Regulating Annealing in Heavy-Tailed Diffusion Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18931","citing_title":"Markov Chain Decoders Overcome the Heavy-Tail Limitations of Lipschitz Generative Models","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ","json":"https://pith.science/pith/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ.json","graph_json":"https://pith.science/api/pith-number/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ/graph.json","events_json":"https://pith.science/api/pith-number/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ/events.json","paper":"https://pith.science/paper/JA6RQZYS"},"agent_actions":{"view_html":"https://pith.science/pith/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ","download_json":"https://pith.science/pith/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ.json","view_paper":"https://pith.science/paper/JA6RQZYS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.01133&json=true","fetch_graph":"https://pith.science/api/pith-number/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ/graph.json","fetch_events":"https://pith.science/api/pith-number/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ/action/storage_attestation","attest_author":"https://pith.science/pith/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ/action/author_attestation","sign_citation":"https://pith.science/pith/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ/action/citation_signature","submit_replication":"https://pith.science/pith/JA6RQZYS2L2ZJ3KDIPY4Q55RMQ/action/replication_record"}},"created_at":"2026-07-05T07:51:36.014623+00:00","updated_at":"2026-07-05T07:51:36.014623+00:00"}