{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:TWUF2IML44NDVJPK6E7U7MAIRS","short_pith_number":"pith:TWUF2IML","schema_version":"1.0","canonical_sha256":"9da85d218be71a3aa5eaf13f4fb0088c92d681aa3c574ea25994254e8d2ab82d","source":{"kind":"arxiv","id":"2311.17901","version":1},"attestation_state":"computed","paper":{"title":"SODA: Bottleneck Diffusion Models for Representation Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Alexander Lerchner, Andrew Jaegle, Andrew K. Lampinen, Daniel Zoran, Drew A. Hudson, Felix Hill, James L. McClelland, Loic Matthey, Mateusz Malinowski","submitted_at":"2023-11-29T18:53:34Z","abstract_excerpt":"We introduce SODA, a self-supervised diffusion model, designed for representation learning. The model incorporates an image encoder, which distills a source view into a compact representation, that, in turn, guides the generation of related novel views. We show that by imposing a tight bottleneck between the encoder and a denoising decoder, and leveraging novel view synthesis as a self-supervised objective, we can turn diffusion models into strong representation learners, capable of capturing visual semantics in an unsupervised manner. To the best of our knowledge, SODA is the first diffusion "},"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":"2311.17901","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-11-29T18:53:34Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"7a2d9a423a70ec8f3e50837432e6340351a6f18135af5def6065d5ebe95dd0b3","abstract_canon_sha256":"579deb241722129c70a85ef4f05823dd89355269d1e7c82a5c0dc3b78c87d176"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:18:20.785093Z","signature_b64":"wphIZkOCCVn4ap9CNNQStZ4GOfkfIQG7fogh4VQ8CqBNJGRnJiXYWkXUOWrU8PfaOA6D7+T+baBdVivVumQWAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9da85d218be71a3aa5eaf13f4fb0088c92d681aa3c574ea25994254e8d2ab82d","last_reissued_at":"2026-07-05T07:18:20.784604Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:18:20.784604Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SODA: Bottleneck Diffusion Models for Representation Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Alexander Lerchner, Andrew Jaegle, Andrew K. Lampinen, Daniel Zoran, Drew A. Hudson, Felix Hill, James L. McClelland, Loic Matthey, Mateusz Malinowski","submitted_at":"2023-11-29T18:53:34Z","abstract_excerpt":"We introduce SODA, a self-supervised diffusion model, designed for representation learning. The model incorporates an image encoder, which distills a source view into a compact representation, that, in turn, guides the generation of related novel views. We show that by imposing a tight bottleneck between the encoder and a denoising decoder, and leveraging novel view synthesis as a self-supervised objective, we can turn diffusion models into strong representation learners, capable of capturing visual semantics in an unsupervised manner. To the best of our knowledge, SODA is the first diffusion "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.17901","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/2311.17901/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":"2311.17901","created_at":"2026-07-05T07:18:20.784655+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.17901v1","created_at":"2026-07-05T07:18:20.784655+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.17901","created_at":"2026-07-05T07:18:20.784655+00:00"},{"alias_kind":"pith_short_12","alias_value":"TWUF2IML44ND","created_at":"2026-07-05T07:18:20.784655+00:00"},{"alias_kind":"pith_short_16","alias_value":"TWUF2IML44NDVJPK","created_at":"2026-07-05T07:18:20.784655+00:00"},{"alias_kind":"pith_short_8","alias_value":"TWUF2IML","created_at":"2026-07-05T07:18:20.784655+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.08376","citing_title":"Bitrate-Controlled Diffusion for Disentangling Motion and Content in Video","ref_index":32,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TWUF2IML44NDVJPK6E7U7MAIRS","json":"https://pith.science/pith/TWUF2IML44NDVJPK6E7U7MAIRS.json","graph_json":"https://pith.science/api/pith-number/TWUF2IML44NDVJPK6E7U7MAIRS/graph.json","events_json":"https://pith.science/api/pith-number/TWUF2IML44NDVJPK6E7U7MAIRS/events.json","paper":"https://pith.science/paper/TWUF2IML"},"agent_actions":{"view_html":"https://pith.science/pith/TWUF2IML44NDVJPK6E7U7MAIRS","download_json":"https://pith.science/pith/TWUF2IML44NDVJPK6E7U7MAIRS.json","view_paper":"https://pith.science/paper/TWUF2IML","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.17901&json=true","fetch_graph":"https://pith.science/api/pith-number/TWUF2IML44NDVJPK6E7U7MAIRS/graph.json","fetch_events":"https://pith.science/api/pith-number/TWUF2IML44NDVJPK6E7U7MAIRS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TWUF2IML44NDVJPK6E7U7MAIRS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TWUF2IML44NDVJPK6E7U7MAIRS/action/storage_attestation","attest_author":"https://pith.science/pith/TWUF2IML44NDVJPK6E7U7MAIRS/action/author_attestation","sign_citation":"https://pith.science/pith/TWUF2IML44NDVJPK6E7U7MAIRS/action/citation_signature","submit_replication":"https://pith.science/pith/TWUF2IML44NDVJPK6E7U7MAIRS/action/replication_record"}},"created_at":"2026-07-05T07:18:20.784655+00:00","updated_at":"2026-07-05T07:18:20.784655+00:00"}