{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:R3SLMFZPKZTVQ4TK6BQZEZBOFF","short_pith_number":"pith:R3SLMFZP","schema_version":"1.0","canonical_sha256":"8ee4b6172f566758726af06192642e294709c64affc6b2e29a61a49f98d983c9","source":{"kind":"arxiv","id":"2312.17172","version":1},"attestation_state":"computed","paper":{"title":"Unified-IO 2: Scaling Autoregressive Multimodal Models with Vision, Language, Audio, and Action","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Aniruddha Kembhavi, Christopher Clark, Derek Hoiem, Jiasen Lu, Ryan Marten, Sangho Lee, Savya Khosla, Zichen Zhang","submitted_at":"2023-12-28T17:57:06Z","abstract_excerpt":"We present Unified-IO 2, the first autoregressive multimodal model that is capable of understanding and generating image, text, audio, and action. To unify different modalities, we tokenize inputs and outputs -- images, text, audio, action, bounding boxes, etc., into a shared semantic space and then process them with a single encoder-decoder transformer model. Since training with such diverse modalities is challenging, we propose various architectural improvements to stabilize model training. We train our model from scratch on a large multimodal pre-training corpus from diverse sources with a "},"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":"2312.17172","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-12-28T17:57:06Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"8b2a2522c0f67d039e0debb2176a8a0c9600c7a117468c61eb092627845bee55","abstract_canon_sha256":"be9a63eda8bcad0dcf2c2f72f96b57100090df26fbb59a301665b42e515e9633"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:28:40.406260Z","signature_b64":"g0T39IgxPHp3aAQobCxMkKKkLpme7yMPSop50x8wlUAw81hOD+FewGL4RjHiXycjEdKrTGkGodI9BLpseVdVAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8ee4b6172f566758726af06192642e294709c64affc6b2e29a61a49f98d983c9","last_reissued_at":"2026-07-05T07:28:40.405805Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:28:40.405805Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Unified-IO 2: Scaling Autoregressive Multimodal Models with Vision, Language, Audio, and Action","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Aniruddha Kembhavi, Christopher Clark, Derek Hoiem, Jiasen Lu, Ryan Marten, Sangho Lee, Savya Khosla, Zichen Zhang","submitted_at":"2023-12-28T17:57:06Z","abstract_excerpt":"We present Unified-IO 2, the first autoregressive multimodal model that is capable of understanding and generating image, text, audio, and action. To unify different modalities, we tokenize inputs and outputs -- images, text, audio, action, bounding boxes, etc., into a shared semantic space and then process them with a single encoder-decoder transformer model. Since training with such diverse modalities is challenging, we propose various architectural improvements to stabilize model training. We train our model from scratch on a large multimodal pre-training corpus from diverse sources with a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.17172","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/2312.17172/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":"2312.17172","created_at":"2026-07-05T07:28:40.405855+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.17172v1","created_at":"2026-07-05T07:28:40.405855+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.17172","created_at":"2026-07-05T07:28:40.405855+00:00"},{"alias_kind":"pith_short_12","alias_value":"R3SLMFZPKZTV","created_at":"2026-07-05T07:28:40.405855+00:00"},{"alias_kind":"pith_short_16","alias_value":"R3SLMFZPKZTVQ4TK","created_at":"2026-07-05T07:28:40.405855+00:00"},{"alias_kind":"pith_short_8","alias_value":"R3SLMFZP","created_at":"2026-07-05T07:28:40.405855+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.18714","citing_title":"Semantic Generative Tuning for Unified Multimodal Models","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2406.09333","citing_title":"Learning Spatial-Preserving Hierarchical Representations for Digital Pathology","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18714","citing_title":"Semantic Generative Tuning for Unified Multimodal Models","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2406.09411","citing_title":"MuirBench: A Comprehensive Benchmark for Robust Multi-image Understanding","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2404.14396","citing_title":"SEED-X: Multimodal Models with Unified Multi-granularity Comprehension and Generation","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2404.12390","citing_title":"BLINK: Multimodal Large Language Models Can See but Not Perceive","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2506.15564","citing_title":"Show-o2: Improved Native Unified Multimodal Models","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2407.07726","citing_title":"PaliGemma: A versatile 3B VLM for transfer","ref_index":86,"is_internal_anchor":false},{"citing_arxiv_id":"2405.09818","citing_title":"Chameleon: Mixed-Modal Early-Fusion Foundation Models","ref_index":20,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/R3SLMFZPKZTVQ4TK6BQZEZBOFF","json":"https://pith.science/pith/R3SLMFZPKZTVQ4TK6BQZEZBOFF.json","graph_json":"https://pith.science/api/pith-number/R3SLMFZPKZTVQ4TK6BQZEZBOFF/graph.json","events_json":"https://pith.science/api/pith-number/R3SLMFZPKZTVQ4TK6BQZEZBOFF/events.json","paper":"https://pith.science/paper/R3SLMFZP"},"agent_actions":{"view_html":"https://pith.science/pith/R3SLMFZPKZTVQ4TK6BQZEZBOFF","download_json":"https://pith.science/pith/R3SLMFZPKZTVQ4TK6BQZEZBOFF.json","view_paper":"https://pith.science/paper/R3SLMFZP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.17172&json=true","fetch_graph":"https://pith.science/api/pith-number/R3SLMFZPKZTVQ4TK6BQZEZBOFF/graph.json","fetch_events":"https://pith.science/api/pith-number/R3SLMFZPKZTVQ4TK6BQZEZBOFF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/R3SLMFZPKZTVQ4TK6BQZEZBOFF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/R3SLMFZPKZTVQ4TK6BQZEZBOFF/action/storage_attestation","attest_author":"https://pith.science/pith/R3SLMFZPKZTVQ4TK6BQZEZBOFF/action/author_attestation","sign_citation":"https://pith.science/pith/R3SLMFZPKZTVQ4TK6BQZEZBOFF/action/citation_signature","submit_replication":"https://pith.science/pith/R3SLMFZPKZTVQ4TK6BQZEZBOFF/action/replication_record"}},"created_at":"2026-07-05T07:28:40.405855+00:00","updated_at":"2026-07-05T07:28:40.405855+00:00"}