{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:R5H7DXU56JNBXNHDN5W6Q3DMRL","short_pith_number":"pith:R5H7DXU5","schema_version":"1.0","canonical_sha256":"8f4ff1de9df25a1bb4e36f6de86c6c8affa9fd855c5548811da710d8e6fde225","source":{"kind":"arxiv","id":"2412.06673","version":1},"attestation_state":"computed","paper":{"title":"ILLUME: Illuminating Your LLMs to See, Draw, and Self-Enhance","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chunwei Wang, Guansong Lu, Hang Xu, Jianhua Han, Junwei Yang, Lu Hou, Runhui Huang, Wei Zhang","submitted_at":"2024-12-09T17:11:50Z","abstract_excerpt":"In this paper, we introduce ILLUME, a unified multimodal large language model (MLLM) that seamlessly integrates multimodal understanding and generation capabilities within a single large language model through a unified next-token prediction formulation. To address the large dataset size typically required for image-text alignment, we propose to enhance data efficiency through the design of a vision tokenizer that incorporates semantic information and a progressive multi-stage training procedure. This approach reduces the dataset size to just 15M for pretraining -- over four times fewer than w"},"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":"2412.06673","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-09T17:11:50Z","cross_cats_sorted":[],"title_canon_sha256":"6f012f516a5ff2c8ec8f3933d4afac5f2c4093a92c352e5401af321fa1035d6b","abstract_canon_sha256":"829f3a2feda957234a6ae875a98019d944cf8e7931788adaa32fb9165cd7444f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:46:36.190398Z","signature_b64":"mRgp85qpPNWv3nnuNETifzwP9RBKRhs4sJqqh01VmEgwIJUR3sV3zMskdR3b8FfoSkb+tofg4I6urN9q1U7/CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8f4ff1de9df25a1bb4e36f6de86c6c8affa9fd855c5548811da710d8e6fde225","last_reissued_at":"2026-07-05T09:46:36.189890Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:46:36.189890Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ILLUME: Illuminating Your LLMs to See, Draw, and Self-Enhance","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chunwei Wang, Guansong Lu, Hang Xu, Jianhua Han, Junwei Yang, Lu Hou, Runhui Huang, Wei Zhang","submitted_at":"2024-12-09T17:11:50Z","abstract_excerpt":"In this paper, we introduce ILLUME, a unified multimodal large language model (MLLM) that seamlessly integrates multimodal understanding and generation capabilities within a single large language model through a unified next-token prediction formulation. To address the large dataset size typically required for image-text alignment, we propose to enhance data efficiency through the design of a vision tokenizer that incorporates semantic information and a progressive multi-stage training procedure. This approach reduces the dataset size to just 15M for pretraining -- over four times fewer than w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.06673","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/2412.06673/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":"2412.06673","created_at":"2026-07-05T09:46:36.189948+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.06673v1","created_at":"2026-07-05T09:46:36.189948+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.06673","created_at":"2026-07-05T09:46:36.189948+00:00"},{"alias_kind":"pith_short_12","alias_value":"R5H7DXU56JNB","created_at":"2026-07-05T09:46:36.189948+00:00"},{"alias_kind":"pith_short_16","alias_value":"R5H7DXU56JNBXNHD","created_at":"2026-07-05T09:46:36.189948+00:00"},{"alias_kind":"pith_short_8","alias_value":"R5H7DXU5","created_at":"2026-07-05T09:46:36.189948+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.17296","citing_title":"Pareto LoRA: Mitigating Modality Imbalance in Unified Multimodal Models via Pareto-Optimal Gradient Integration","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13289","citing_title":"HYDRA-X: Native Unified Multimodal Models with Holistic Visual Tokenizers","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2505.23606","citing_title":"Muddit: Liberating Generation Beyond Text-to-Image with a Unified Discrete Diffusion Model","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2509.21912","citing_title":"Discrete Guidance Matching: Exact Guidance for Discrete Flow Matching","ref_index":77,"is_internal_anchor":false},{"citing_arxiv_id":"2505.05472","citing_title":"Mogao: An Omni Foundation Model for Interleaved Multi-Modal Generation","ref_index":79,"is_internal_anchor":false},{"citing_arxiv_id":"2601.15507","citing_title":"A Unified and Controllable Framework for Layered Image Generation with Visual Effects","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2506.15564","citing_title":"Show-o2: Improved Native Unified Multimodal Models","ref_index":111,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10564","citing_title":"DeepSight: Long-Horizon World Modeling via Latent States Prediction for End-to-End Autonomous Driving","ref_index":120,"is_internal_anchor":false},{"citing_arxiv_id":"2501.17811","citing_title":"Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2505.14683","citing_title":"Emerging Properties in Unified Multimodal Pretraining","ref_index":77,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/R5H7DXU56JNBXNHDN5W6Q3DMRL","json":"https://pith.science/pith/R5H7DXU56JNBXNHDN5W6Q3DMRL.json","graph_json":"https://pith.science/api/pith-number/R5H7DXU56JNBXNHDN5W6Q3DMRL/graph.json","events_json":"https://pith.science/api/pith-number/R5H7DXU56JNBXNHDN5W6Q3DMRL/events.json","paper":"https://pith.science/paper/R5H7DXU5"},"agent_actions":{"view_html":"https://pith.science/pith/R5H7DXU56JNBXNHDN5W6Q3DMRL","download_json":"https://pith.science/pith/R5H7DXU56JNBXNHDN5W6Q3DMRL.json","view_paper":"https://pith.science/paper/R5H7DXU5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.06673&json=true","fetch_graph":"https://pith.science/api/pith-number/R5H7DXU56JNBXNHDN5W6Q3DMRL/graph.json","fetch_events":"https://pith.science/api/pith-number/R5H7DXU56JNBXNHDN5W6Q3DMRL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/R5H7DXU56JNBXNHDN5W6Q3DMRL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/R5H7DXU56JNBXNHDN5W6Q3DMRL/action/storage_attestation","attest_author":"https://pith.science/pith/R5H7DXU56JNBXNHDN5W6Q3DMRL/action/author_attestation","sign_citation":"https://pith.science/pith/R5H7DXU56JNBXNHDN5W6Q3DMRL/action/citation_signature","submit_replication":"https://pith.science/pith/R5H7DXU56JNBXNHDN5W6Q3DMRL/action/replication_record"}},"created_at":"2026-07-05T09:46:36.189948+00:00","updated_at":"2026-07-05T09:46:36.189948+00:00"}