{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:ZTM2DKCUVLMU4KWIXTYMDSIGBO","short_pith_number":"pith:ZTM2DKCU","schema_version":"1.0","canonical_sha256":"ccd9a1a854aad94e2ac8bcf0c1c9060b9b6066a6dda485f75eb0acf51716824e","source":{"kind":"arxiv","id":"2301.13622","version":2},"attestation_state":"computed","paper":{"title":"Learning Data Representations with Joint Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jakub M. Tomczak, Kamil Deja, Tomasz Trzcinski","submitted_at":"2023-01-31T13:29:19Z","abstract_excerpt":"Joint machine learning models that allow synthesizing and classifying data often offer uneven performance between those tasks or are unstable to train. In this work, we depart from a set of empirical observations that indicate the usefulness of internal representations built by contemporary deep diffusion-based generative models not only for generating but also predicting. We then propose to extend the vanilla diffusion model with a classifier that allows for stable joint end-to-end training with shared parameterization between those objectives. The resulting joint diffusion model outperforms "},"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":"2301.13622","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-01-31T13:29:19Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"334bec1842db6e33511f9f08e76d26c88963f40158899577dd91fde623ee505a","abstract_canon_sha256":"19abd16146c3dd88232a47005f685462514718cf2770d652274a3680f27c5bcb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:58:19.494075Z","signature_b64":"jjvW0vhbZJ0g24bGdkxmlYC2vt9Wpkn24wpHVT7Cyf6kfIKcPRqNH050Rs5dx+ahm3P20KT4puuO7loYG6bIDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ccd9a1a854aad94e2ac8bcf0c1c9060b9b6066a6dda485f75eb0acf51716824e","last_reissued_at":"2026-07-05T05:58:19.493651Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:58:19.493651Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Data Representations with Joint Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jakub M. Tomczak, Kamil Deja, Tomasz Trzcinski","submitted_at":"2023-01-31T13:29:19Z","abstract_excerpt":"Joint machine learning models that allow synthesizing and classifying data often offer uneven performance between those tasks or are unstable to train. In this work, we depart from a set of empirical observations that indicate the usefulness of internal representations built by contemporary deep diffusion-based generative models not only for generating but also predicting. We then propose to extend the vanilla diffusion model with a classifier that allows for stable joint end-to-end training with shared parameterization between those objectives. The resulting joint diffusion model outperforms "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2301.13622","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/2301.13622/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":"2301.13622","created_at":"2026-07-05T05:58:19.493713+00:00"},{"alias_kind":"arxiv_version","alias_value":"2301.13622v2","created_at":"2026-07-05T05:58:19.493713+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2301.13622","created_at":"2026-07-05T05:58:19.493713+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZTM2DKCUVLMU","created_at":"2026-07-05T05:58:19.493713+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZTM2DKCUVLMU4KWI","created_at":"2026-07-05T05:58:19.493713+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZTM2DKCU","created_at":"2026-07-05T05:58:19.493713+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.09434","citing_title":"Mediffusion: Joint Diffusion for Self-Explainable Semi-Supervised Classification and Medical Image Generation","ref_index":23,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZTM2DKCUVLMU4KWIXTYMDSIGBO","json":"https://pith.science/pith/ZTM2DKCUVLMU4KWIXTYMDSIGBO.json","graph_json":"https://pith.science/api/pith-number/ZTM2DKCUVLMU4KWIXTYMDSIGBO/graph.json","events_json":"https://pith.science/api/pith-number/ZTM2DKCUVLMU4KWIXTYMDSIGBO/events.json","paper":"https://pith.science/paper/ZTM2DKCU"},"agent_actions":{"view_html":"https://pith.science/pith/ZTM2DKCUVLMU4KWIXTYMDSIGBO","download_json":"https://pith.science/pith/ZTM2DKCUVLMU4KWIXTYMDSIGBO.json","view_paper":"https://pith.science/paper/ZTM2DKCU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2301.13622&json=true","fetch_graph":"https://pith.science/api/pith-number/ZTM2DKCUVLMU4KWIXTYMDSIGBO/graph.json","fetch_events":"https://pith.science/api/pith-number/ZTM2DKCUVLMU4KWIXTYMDSIGBO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZTM2DKCUVLMU4KWIXTYMDSIGBO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZTM2DKCUVLMU4KWIXTYMDSIGBO/action/storage_attestation","attest_author":"https://pith.science/pith/ZTM2DKCUVLMU4KWIXTYMDSIGBO/action/author_attestation","sign_citation":"https://pith.science/pith/ZTM2DKCUVLMU4KWIXTYMDSIGBO/action/citation_signature","submit_replication":"https://pith.science/pith/ZTM2DKCUVLMU4KWIXTYMDSIGBO/action/replication_record"}},"created_at":"2026-07-05T05:58:19.493713+00:00","updated_at":"2026-07-05T05:58:19.493713+00:00"}