{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:Y3E4AMUMOU264Q4UTV5SZKOHAQ","short_pith_number":"pith:Y3E4AMUM","schema_version":"1.0","canonical_sha256":"c6c9c0328c7535ee43949d7b2ca9c7040c971dd13cfd08020f65f6634a4d989c","source":{"kind":"arxiv","id":"2312.06647","version":1},"attestation_state":"computed","paper":{"title":"4M: Massively Multimodal Masked Modeling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Afshin Dehghan, Amir Zamir, David Mizrahi, Mingfei Gao, O\\u{g}uzhan Fatih Kar, Roman Bachmann, Teresa Yeo","submitted_at":"2023-12-11T18:57:35Z","abstract_excerpt":"Current machine learning models for vision are often highly specialized and limited to a single modality and task. In contrast, recent large language models exhibit a wide range of capabilities, hinting at a possibility for similarly versatile models in computer vision. In this paper, we take a step in this direction and propose a multimodal training scheme called 4M. It consists of training a single unified Transformer encoder-decoder using a masked modeling objective across a wide range of input/output modalities - including text, images, geometric, and semantic modalities, as well as neural"},"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":"2312.06647","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-12-11T18:57:35Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"cd7d0914d6c1db5e9594b51755534dac06ddae74482f8951d6a69c57b6fbaadf","abstract_canon_sha256":"f2be6e54540d596d04ad7231d4ea0092049b76e934606ebb5a64db190e0cc484"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:22:48.329361Z","signature_b64":"0n5tgrFNJR3UXdXY/ToiiJxVburOviBzp2pJj3Dt/5yWvnU6n/3BiiIC3vBBSkyXnmYAmqdlOEaev/wXWxzYAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c6c9c0328c7535ee43949d7b2ca9c7040c971dd13cfd08020f65f6634a4d989c","last_reissued_at":"2026-07-05T07:22:48.328867Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:22:48.328867Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"4M: Massively Multimodal Masked Modeling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Afshin Dehghan, Amir Zamir, David Mizrahi, Mingfei Gao, O\\u{g}uzhan Fatih Kar, Roman Bachmann, Teresa Yeo","submitted_at":"2023-12-11T18:57:35Z","abstract_excerpt":"Current machine learning models for vision are often highly specialized and limited to a single modality and task. In contrast, recent large language models exhibit a wide range of capabilities, hinting at a possibility for similarly versatile models in computer vision. In this paper, we take a step in this direction and propose a multimodal training scheme called 4M. It consists of training a single unified Transformer encoder-decoder using a masked modeling objective across a wide range of input/output modalities - including text, images, geometric, and semantic modalities, as well as neural"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.06647","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/2312.06647/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":"2312.06647","created_at":"2026-07-05T07:22:48.328929+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.06647v1","created_at":"2026-07-05T07:22:48.328929+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.06647","created_at":"2026-07-05T07:22:48.328929+00:00"},{"alias_kind":"pith_short_12","alias_value":"Y3E4AMUMOU26","created_at":"2026-07-05T07:22:48.328929+00:00"},{"alias_kind":"pith_short_16","alias_value":"Y3E4AMUMOU264Q4U","created_at":"2026-07-05T07:22:48.328929+00:00"},{"alias_kind":"pith_short_8","alias_value":"Y3E4AMUM","created_at":"2026-07-05T07:22:48.328929+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.18703","citing_title":"Contextualizing Biological Language Models across Modalities via Logit-Space Contrastive Alignment","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31215","citing_title":"Fixed-Point Masked Generative Modeling","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21921","citing_title":"Context Unrolling in Omni Models","ref_index":33,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Y3E4AMUMOU264Q4UTV5SZKOHAQ","json":"https://pith.science/pith/Y3E4AMUMOU264Q4UTV5SZKOHAQ.json","graph_json":"https://pith.science/api/pith-number/Y3E4AMUMOU264Q4UTV5SZKOHAQ/graph.json","events_json":"https://pith.science/api/pith-number/Y3E4AMUMOU264Q4UTV5SZKOHAQ/events.json","paper":"https://pith.science/paper/Y3E4AMUM"},"agent_actions":{"view_html":"https://pith.science/pith/Y3E4AMUMOU264Q4UTV5SZKOHAQ","download_json":"https://pith.science/pith/Y3E4AMUMOU264Q4UTV5SZKOHAQ.json","view_paper":"https://pith.science/paper/Y3E4AMUM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.06647&json=true","fetch_graph":"https://pith.science/api/pith-number/Y3E4AMUMOU264Q4UTV5SZKOHAQ/graph.json","fetch_events":"https://pith.science/api/pith-number/Y3E4AMUMOU264Q4UTV5SZKOHAQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Y3E4AMUMOU264Q4UTV5SZKOHAQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Y3E4AMUMOU264Q4UTV5SZKOHAQ/action/storage_attestation","attest_author":"https://pith.science/pith/Y3E4AMUMOU264Q4UTV5SZKOHAQ/action/author_attestation","sign_citation":"https://pith.science/pith/Y3E4AMUMOU264Q4UTV5SZKOHAQ/action/citation_signature","submit_replication":"https://pith.science/pith/Y3E4AMUMOU264Q4UTV5SZKOHAQ/action/replication_record"}},"created_at":"2026-07-05T07:22:48.328929+00:00","updated_at":"2026-07-05T07:22:48.328929+00:00"}