{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:E7FT6OWI45SQB72EDKS7UFCBLU","short_pith_number":"pith:E7FT6OWI","schema_version":"1.0","canonical_sha256":"27cb3f3ac8e76500ff441aa5fa14415d2100d8cefca6094d0c9260a66f64ed44","source":{"kind":"arxiv","id":"2406.17688","version":1},"attestation_state":"computed","paper":{"title":"Unified Auto-Encoding with Masked Diffusion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Amy Zhang, Manan Tomar, Philippe Hansen-Estruch, Sriram Vishwanath","submitted_at":"2024-06-25T16:24:34Z","abstract_excerpt":"At the core of both successful generative and self-supervised representation learning models there is a reconstruction objective that incorporates some form of image corruption. Diffusion models implement this approach through a scheduled Gaussian corruption process, while masked auto-encoder models do so by masking patches of the image. Despite their different approaches, the underlying similarity in their methodologies suggests a promising avenue for an auto-encoder capable of both de-noising tasks. We propose a unified self-supervised objective, dubbed Unified Masked Diffusion (UMD), that c"},"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":"2406.17688","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-06-25T16:24:34Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5c4e2988e60e95b0d3e33678ed83fc3bb9673246481b33b11408ed44b7e674a5","abstract_canon_sha256":"9b4095890b45850e5e5b8d665fe0c9717f479c8fcfe6b07a7921aed06cd9a814"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:36:38.693229Z","signature_b64":"rXaubkbYc4osZJIK6BBqRFEB6a4JTMs0Gu90n5iqZV/Oy34vrzvZkILX2PjmA0ISzAWbtVJvMt0p3H+hgT0XDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"27cb3f3ac8e76500ff441aa5fa14415d2100d8cefca6094d0c9260a66f64ed44","last_reissued_at":"2026-07-05T08:36:38.692647Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:36:38.692647Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Unified Auto-Encoding with Masked Diffusion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Amy Zhang, Manan Tomar, Philippe Hansen-Estruch, Sriram Vishwanath","submitted_at":"2024-06-25T16:24:34Z","abstract_excerpt":"At the core of both successful generative and self-supervised representation learning models there is a reconstruction objective that incorporates some form of image corruption. Diffusion models implement this approach through a scheduled Gaussian corruption process, while masked auto-encoder models do so by masking patches of the image. Despite their different approaches, the underlying similarity in their methodologies suggests a promising avenue for an auto-encoder capable of both de-noising tasks. We propose a unified self-supervised objective, dubbed Unified Masked Diffusion (UMD), that c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.17688","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/2406.17688/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":"2406.17688","created_at":"2026-07-05T08:36:38.692727+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.17688v1","created_at":"2026-07-05T08:36:38.692727+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.17688","created_at":"2026-07-05T08:36:38.692727+00:00"},{"alias_kind":"pith_short_12","alias_value":"E7FT6OWI45SQ","created_at":"2026-07-05T08:36:38.692727+00:00"},{"alias_kind":"pith_short_16","alias_value":"E7FT6OWI45SQB72E","created_at":"2026-07-05T08:36:38.692727+00:00"},{"alias_kind":"pith_short_8","alias_value":"E7FT6OWI","created_at":"2026-07-05T08:36:38.692727+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.07466","citing_title":"Less is More: Masking Elements in Image Condition Features Avoids Content Leakages in Style Transfer Diffusion Models","ref_index":10,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/E7FT6OWI45SQB72EDKS7UFCBLU","json":"https://pith.science/pith/E7FT6OWI45SQB72EDKS7UFCBLU.json","graph_json":"https://pith.science/api/pith-number/E7FT6OWI45SQB72EDKS7UFCBLU/graph.json","events_json":"https://pith.science/api/pith-number/E7FT6OWI45SQB72EDKS7UFCBLU/events.json","paper":"https://pith.science/paper/E7FT6OWI"},"agent_actions":{"view_html":"https://pith.science/pith/E7FT6OWI45SQB72EDKS7UFCBLU","download_json":"https://pith.science/pith/E7FT6OWI45SQB72EDKS7UFCBLU.json","view_paper":"https://pith.science/paper/E7FT6OWI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.17688&json=true","fetch_graph":"https://pith.science/api/pith-number/E7FT6OWI45SQB72EDKS7UFCBLU/graph.json","fetch_events":"https://pith.science/api/pith-number/E7FT6OWI45SQB72EDKS7UFCBLU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/E7FT6OWI45SQB72EDKS7UFCBLU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/E7FT6OWI45SQB72EDKS7UFCBLU/action/storage_attestation","attest_author":"https://pith.science/pith/E7FT6OWI45SQB72EDKS7UFCBLU/action/author_attestation","sign_citation":"https://pith.science/pith/E7FT6OWI45SQB72EDKS7UFCBLU/action/citation_signature","submit_replication":"https://pith.science/pith/E7FT6OWI45SQB72EDKS7UFCBLU/action/replication_record"}},"created_at":"2026-07-05T08:36:38.692727+00:00","updated_at":"2026-07-05T08:36:38.692727+00:00"}