{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:R6JZZFQV5X4AOBUOW2B6PXSM57","short_pith_number":"pith:R6JZZFQV","schema_version":"1.0","canonical_sha256":"8f939c9615edf807068eb683e7de4cefccd4a731c4c66806c3cb2b1151b9a965","source":{"kind":"arxiv","id":"2211.09117","version":2},"attestation_state":"computed","paper":{"title":"MAGE: MAsked Generative Encoder to Unify Representation Learning and Image Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dilip Krishnan, Dina Katabi, Han Zhang, Huiwen Chang, Shlok Kumar Mishra, Tianhong Li","submitted_at":"2022-11-16T18:59:02Z","abstract_excerpt":"Generative modeling and representation learning are two key tasks in computer vision. However, these models are typically trained independently, which ignores the potential for each task to help the other, and leads to training and model maintenance overheads. In this work, we propose MAsked Generative Encoder (MAGE), the first framework to unify SOTA image generation and self-supervised representation learning. Our key insight is that using variable masking ratios in masked image modeling pre-training can allow generative training (very high masking ratio) and representation learning (lower m"},"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":"2211.09117","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2022-11-16T18:59:02Z","cross_cats_sorted":[],"title_canon_sha256":"54fa3a425f1edf5b81a11e660286fbc38f41adc4ad6985c7b58578adefedcddb","abstract_canon_sha256":"4be4bbd0b2c3c027687ef6969a7d656db5901c15d0bbc9625b1cf79e5d0a29b5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:26:17.339890Z","signature_b64":"gmTlATn5hJs/WnQlpconTmFcEqKzlON/9xz0YNM1T+QxRY4/4305f8BONtOlv1jh13ZeVSzhw8FHb12PSD64AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8f939c9615edf807068eb683e7de4cefccd4a731c4c66806c3cb2b1151b9a965","last_reissued_at":"2026-07-05T06:26:17.339449Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:26:17.339449Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MAGE: MAsked Generative Encoder to Unify Representation Learning and Image Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dilip Krishnan, Dina Katabi, Han Zhang, Huiwen Chang, Shlok Kumar Mishra, Tianhong Li","submitted_at":"2022-11-16T18:59:02Z","abstract_excerpt":"Generative modeling and representation learning are two key tasks in computer vision. However, these models are typically trained independently, which ignores the potential for each task to help the other, and leads to training and model maintenance overheads. In this work, we propose MAsked Generative Encoder (MAGE), the first framework to unify SOTA image generation and self-supervised representation learning. Our key insight is that using variable masking ratios in masked image modeling pre-training can allow generative training (very high masking ratio) and representation learning (lower m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.09117","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/2211.09117/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":"2211.09117","created_at":"2026-07-05T06:26:17.339509+00:00"},{"alias_kind":"arxiv_version","alias_value":"2211.09117v2","created_at":"2026-07-05T06:26:17.339509+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.09117","created_at":"2026-07-05T06:26:17.339509+00:00"},{"alias_kind":"pith_short_12","alias_value":"R6JZZFQV5X4A","created_at":"2026-07-05T06:26:17.339509+00:00"},{"alias_kind":"pith_short_16","alias_value":"R6JZZFQV5X4AOBUO","created_at":"2026-07-05T06:26:17.339509+00:00"},{"alias_kind":"pith_short_8","alias_value":"R6JZZFQV","created_at":"2026-07-05T06:26:17.339509+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.21022","citing_title":"Instella-T2I: Pushing the Limits of 1D Discrete Latent Space Image Generation","ref_index":14,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/R6JZZFQV5X4AOBUOW2B6PXSM57","json":"https://pith.science/pith/R6JZZFQV5X4AOBUOW2B6PXSM57.json","graph_json":"https://pith.science/api/pith-number/R6JZZFQV5X4AOBUOW2B6PXSM57/graph.json","events_json":"https://pith.science/api/pith-number/R6JZZFQV5X4AOBUOW2B6PXSM57/events.json","paper":"https://pith.science/paper/R6JZZFQV"},"agent_actions":{"view_html":"https://pith.science/pith/R6JZZFQV5X4AOBUOW2B6PXSM57","download_json":"https://pith.science/pith/R6JZZFQV5X4AOBUOW2B6PXSM57.json","view_paper":"https://pith.science/paper/R6JZZFQV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2211.09117&json=true","fetch_graph":"https://pith.science/api/pith-number/R6JZZFQV5X4AOBUOW2B6PXSM57/graph.json","fetch_events":"https://pith.science/api/pith-number/R6JZZFQV5X4AOBUOW2B6PXSM57/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/R6JZZFQV5X4AOBUOW2B6PXSM57/action/timestamp_anchor","attest_storage":"https://pith.science/pith/R6JZZFQV5X4AOBUOW2B6PXSM57/action/storage_attestation","attest_author":"https://pith.science/pith/R6JZZFQV5X4AOBUOW2B6PXSM57/action/author_attestation","sign_citation":"https://pith.science/pith/R6JZZFQV5X4AOBUOW2B6PXSM57/action/citation_signature","submit_replication":"https://pith.science/pith/R6JZZFQV5X4AOBUOW2B6PXSM57/action/replication_record"}},"created_at":"2026-07-05T06:26:17.339509+00:00","updated_at":"2026-07-05T06:26:17.339509+00:00"}