{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:CE2A5ZGJ62C53PNEDIDFOWRUDY","short_pith_number":"pith:CE2A5ZGJ","schema_version":"1.0","canonical_sha256":"11340ee4c9f685ddbda41a06575a341e1e1a5ca0d38a4401b2b9ddaa9d1535e4","source":{"kind":"arxiv","id":"2305.17216","version":3},"attestation_state":"computed","paper":{"title":"Generating Images with Multimodal Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.CL","authors_text":"Daniel Fried, Jing Yu Koh, Ruslan Salakhutdinov","submitted_at":"2023-05-26T19:22:03Z","abstract_excerpt":"We propose a method to fuse frozen text-only large language models (LLMs) with pre-trained image encoder and decoder models, by mapping between their embedding spaces. Our model demonstrates a wide suite of multimodal capabilities: image retrieval, novel image generation, and multimodal dialogue. Ours is the first approach capable of conditioning on arbitrarily interleaved image and text inputs to generate coherent image (and text) outputs. To achieve strong performance on image generation, we propose an efficient mapping network to ground the LLM to an off-the-shelf text-to-image generation 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":"2305.17216","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-26T19:22:03Z","cross_cats_sorted":["cs.CV","cs.LG"],"title_canon_sha256":"67f854b6632fd3e9d43740caeb0ff1daf92816cef9996a356d457144c387217a","abstract_canon_sha256":"e07b814059a5f76edbd050eea5b2293d3edc99ae19f44ba73dd0a6854ea68bc5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:00:24.172842Z","signature_b64":"Cb5p0CV+10kfK7J3AIlYBrILo1D0HrSLc1r+NSXhU7CiOoJx4b0vQ6yeb9VpwNi39oxHesLohBXjI93H4hkWBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"11340ee4c9f685ddbda41a06575a341e1e1a5ca0d38a4401b2b9ddaa9d1535e4","last_reissued_at":"2026-07-05T07:00:24.172431Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:00:24.172431Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generating Images with Multimodal Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.CL","authors_text":"Daniel Fried, Jing Yu Koh, Ruslan Salakhutdinov","submitted_at":"2023-05-26T19:22:03Z","abstract_excerpt":"We propose a method to fuse frozen text-only large language models (LLMs) with pre-trained image encoder and decoder models, by mapping between their embedding spaces. Our model demonstrates a wide suite of multimodal capabilities: image retrieval, novel image generation, and multimodal dialogue. Ours is the first approach capable of conditioning on arbitrarily interleaved image and text inputs to generate coherent image (and text) outputs. To achieve strong performance on image generation, we propose an efficient mapping network to ground the LLM to an off-the-shelf text-to-image generation m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.17216","kind":"arxiv","version":3},"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/2305.17216/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":"2305.17216","created_at":"2026-07-05T07:00:24.172485+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.17216v3","created_at":"2026-07-05T07:00:24.172485+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.17216","created_at":"2026-07-05T07:00:24.172485+00:00"},{"alias_kind":"pith_short_12","alias_value":"CE2A5ZGJ62C5","created_at":"2026-07-05T07:00:24.172485+00:00"},{"alias_kind":"pith_short_16","alias_value":"CE2A5ZGJ62C53PNE","created_at":"2026-07-05T07:00:24.172485+00:00"},{"alias_kind":"pith_short_8","alias_value":"CE2A5ZGJ","created_at":"2026-07-05T07:00:24.172485+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2308.05374","citing_title":"Trustworthy LLMs: a Survey and Guideline for Evaluating Large Language Models' Alignment","ref_index":140,"is_internal_anchor":false},{"citing_arxiv_id":"2403.09611","citing_title":"MM1: Methods, Analysis & Insights from Multimodal LLM Pre-training","ref_index":54,"is_internal_anchor":false},{"citing_arxiv_id":"2307.16125","citing_title":"SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension","ref_index":21,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CE2A5ZGJ62C53PNEDIDFOWRUDY","json":"https://pith.science/pith/CE2A5ZGJ62C53PNEDIDFOWRUDY.json","graph_json":"https://pith.science/api/pith-number/CE2A5ZGJ62C53PNEDIDFOWRUDY/graph.json","events_json":"https://pith.science/api/pith-number/CE2A5ZGJ62C53PNEDIDFOWRUDY/events.json","paper":"https://pith.science/paper/CE2A5ZGJ"},"agent_actions":{"view_html":"https://pith.science/pith/CE2A5ZGJ62C53PNEDIDFOWRUDY","download_json":"https://pith.science/pith/CE2A5ZGJ62C53PNEDIDFOWRUDY.json","view_paper":"https://pith.science/paper/CE2A5ZGJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.17216&json=true","fetch_graph":"https://pith.science/api/pith-number/CE2A5ZGJ62C53PNEDIDFOWRUDY/graph.json","fetch_events":"https://pith.science/api/pith-number/CE2A5ZGJ62C53PNEDIDFOWRUDY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CE2A5ZGJ62C53PNEDIDFOWRUDY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CE2A5ZGJ62C53PNEDIDFOWRUDY/action/storage_attestation","attest_author":"https://pith.science/pith/CE2A5ZGJ62C53PNEDIDFOWRUDY/action/author_attestation","sign_citation":"https://pith.science/pith/CE2A5ZGJ62C53PNEDIDFOWRUDY/action/citation_signature","submit_replication":"https://pith.science/pith/CE2A5ZGJ62C53PNEDIDFOWRUDY/action/replication_record"}},"created_at":"2026-07-05T07:00:24.172485+00:00","updated_at":"2026-07-05T07:00:24.172485+00:00"}