{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:A5YWT5NPNYXLFFO6RCZBDF3ZZW","short_pith_number":"pith:A5YWT5NP","schema_version":"1.0","canonical_sha256":"077169f5af6e2eb295de88b2119779cd86902265d8155d7d249a5cdc30022e2e","source":{"kind":"arxiv","id":"2505.15809","version":2},"attestation_state":"computed","paper":{"title":"MMaDA: Multimodal Large Diffusion Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A single diffusion architecture unifies text reasoning, multimodal understanding, and image generation without modality-specific parts.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bowen Li, Ke Shen, Ling Yang, Mengdi Wang, Xinchen Zhang, Ye Tian, Yunhai Tong","submitted_at":"2025-05-21T17:59:05Z","abstract_excerpt":"We introduce MMaDA, a novel class of multimodal diffusion foundation models designed to achieve superior performance across diverse domains such as textual reasoning, multimodal understanding, and text-to-image generation. The approach is distinguished by three key innovations: (i) MMaDA adopts a unified diffusion architecture with a shared probabilistic formulation and a modality-agnostic design, eliminating the need for modality-specific components. This architecture ensures seamless integration and processing across different data types. (ii) We implement a mixed long chain-of-thought (CoT)"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2505.15809","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-05-21T17:59:05Z","cross_cats_sorted":[],"title_canon_sha256":"9c0d94bf569c6f3f39389aa289310f824ba48a152bb9e854e82bacb938cf3058","abstract_canon_sha256":"9915f21e3ed164bef883cebc22d229638a7cf8d206b66e5fa1e279444700fc8d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:52.312361Z","signature_b64":"YKimGEKH+ucaW8Lxe9DmkMQpn2JrEwXxe4RK4UfKgXaEnPOimhr2Gpm9BeaM08KkMan6+YcepR76c4QK48EaDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"077169f5af6e2eb295de88b2119779cd86902265d8155d7d249a5cdc30022e2e","last_reissued_at":"2026-05-17T23:38:52.311878Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:52.311878Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MMaDA: Multimodal Large Diffusion Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A single diffusion architecture unifies text reasoning, multimodal understanding, and image generation without modality-specific parts.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bowen Li, Ke Shen, Ling Yang, Mengdi Wang, Xinchen Zhang, Ye Tian, Yunhai Tong","submitted_at":"2025-05-21T17:59:05Z","abstract_excerpt":"We introduce MMaDA, a novel class of multimodal diffusion foundation models designed to achieve superior performance across diverse domains such as textual reasoning, multimodal understanding, and text-to-image generation. The approach is distinguished by three key innovations: (i) MMaDA adopts a unified diffusion architecture with a shared probabilistic formulation and a modality-agnostic design, eliminating the need for modality-specific components. This architecture ensures seamless integration and processing across different data types. (ii) We implement a mixed long chain-of-thought (CoT)"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"MMaDA-8B exhibits strong generalization capabilities as a unified multimodal foundation model. It surpasses powerful models like LLaMA-3-7B and Qwen2-7B in textual reasoning, outperforms Show-o and SEED-X in multimodal understanding, and excels over SDXL and Janus in text-to-image generation.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The shared probabilistic formulation and modality-agnostic design in the unified diffusion architecture is sufficient to seamlessly integrate and process different data types without modality-specific components.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"MMaDA is a unified multimodal diffusion model using mixed chain-of-thought fine-tuning and a new UniGRPO reinforcement learning algorithm that outperforms specialized models in reasoning, understanding, and text-to-image tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A single diffusion architecture unifies text reasoning, multimodal understanding, and image generation without modality-specific parts.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a88e133fea8857fa0f30815db5e92bd269a2c871019fca0e4649f449dc7b88df"},"source":{"id":"2505.15809","kind":"arxiv","version":2},"verdict":{"id":"9895232e-4f05-421d-b7c2-9de72efa4bbb","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T14:47:07.216120Z","strongest_claim":"MMaDA-8B exhibits strong generalization capabilities as a unified multimodal foundation model. It surpasses powerful models like LLaMA-3-7B and Qwen2-7B in textual reasoning, outperforms Show-o and SEED-X in multimodal understanding, and excels over SDXL and Janus in text-to-image generation.","one_line_summary":"MMaDA is a unified multimodal diffusion model using mixed chain-of-thought fine-tuning and a new UniGRPO reinforcement learning algorithm that outperforms specialized models in reasoning, understanding, and text-to-image tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The shared probabilistic formulation and modality-agnostic design in the unified diffusion architecture is sufficient to seamlessly integrate and process different data types without modality-specific components.","pith_extraction_headline":"A single diffusion architecture unifies text reasoning, multimodal understanding, and image generation without modality-specific parts."},"references":{"count":126,"sample":[{"doi":"","year":2018,"title":"Improving language understanding by generative pre-training","work_id":"72bdd905-4f91-46af-883a-4c2849c99ffd","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Language models are few-shot learners","work_id":"97345cea-ff46-4103-a836-60d34717536c","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"OpenAI o1 System Card","work_id":"68d3c334-0fc9-49e3-b7b0-a69afae933e2","ref_index":3,"cited_arxiv_id":"2412.16720","is_internal_anchor":true},{"doi":"","year":2023,"title":"Vl-gpt: A generative pre-trained transformer for vision and language understanding and generation","work_id":"2891128c-bc6e-4651-be44-a7b8f6f3ad91","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Emu: Generative Pretraining in Multimodality","work_id":"d20ff802-68ef-4783-943f-7bf17acb6a68","ref_index":6,"cited_arxiv_id":"2307.05222","is_internal_anchor":true}],"resolved_work":126,"snapshot_sha256":"5250db472d273842a04ea9d80139cfb7ea5056430e1c598a99862e7d69363eed","internal_anchors":32},"formal_canon":{"evidence_count":2,"snapshot_sha256":"fb2a8088bfdccd40909723b92584d7d46db4a9731f0e15ca3a443baeb5f5803a"},"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":"2505.15809","created_at":"2026-05-17T23:38:52.311957+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.15809v2","created_at":"2026-05-17T23:38:52.311957+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.15809","created_at":"2026-05-17T23:38:52.311957+00:00"},{"alias_kind":"pith_short_12","alias_value":"A5YWT5NPNYXL","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"A5YWT5NPNYXLFFO6","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"A5YWT5NP","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":31,"internal_anchor_count":31,"sample":[{"citing_arxiv_id":"2602.18176","citing_title":"Improving Sampling for Masked Diffusion Models via Information Gain","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2605.23163","citing_title":"Fast-dDrive: Efficient Block-Diffusion VLM for Autonomous Driving","ref_index":21,"is_internal_anchor":true},{"citing_arxiv_id":"2604.10784","citing_title":"TorchUMM: A Unified Multimodal Model Codebase for Evaluation, Analysis, and Post-training","ref_index":29,"is_internal_anchor":true},{"citing_arxiv_id":"2604.24763","citing_title":"Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation","ref_index":48,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16941","citing_title":"Roll Out and Roll Back: Diffusion LLMs are Their Own Efficiency Teachers","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2604.08302","citing_title":"DMax: Aggressive Parallel Decoding for dLLMs","ref_index":90,"is_internal_anchor":true},{"citing_arxiv_id":"2605.14531","citing_title":"Language Generation as Optimal Control: Closed-Loop Diffusion in Latent Control Space","ref_index":45,"is_internal_anchor":true},{"citing_arxiv_id":"2509.20863","citing_title":"GIFT: Guided Importance-Aware Fine-Tuning for Diffusion Language Models","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2509.21912","citing_title":"Discrete Guidance Matching: Exact Guidance for Discrete Flow Matching","ref_index":83,"is_internal_anchor":true},{"citing_arxiv_id":"2512.14067","citing_title":"Efficient-DLM: From Autoregressive to Diffusion Language Models, and Beyond in Speed","ref_index":38,"is_internal_anchor":true},{"citing_arxiv_id":"2512.19433","citing_title":"dMLLM-TTS: Self-Verified and Efficient Test-Time Scaling for Diffusion Multi-Modal Large Language Models","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2603.12554","citing_title":"Reinforcement Learning for Diffusion LLMs with Entropy-Guided Step Selection and Stepwise Advantages","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2605.14531","citing_title":"Language Generation as Optimal Control: Closed-Loop Diffusion in Latent Control Space","ref_index":45,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11726","citing_title":"Block-R1: Rethinking the Role of Block Size in Multi-domain Reinforcement Learning for Diffusion Large Language Models","ref_index":68,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11726","citing_title":"Block-R1: Rethinking the Role of Block Size in Multi-domain Reinforcement Learning for Diffusion Large Language Models","ref_index":68,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11400","citing_title":"UniPath: Adaptive Coordination of Understanding and Generation for Unified Multimodal Reasoning","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2506.15564","citing_title":"Show-o2: Improved Native Unified Multimodal Models","ref_index":135,"is_internal_anchor":true},{"citing_arxiv_id":"2512.13030","citing_title":"Motus: A Unified Latent Action World Model","ref_index":52,"is_internal_anchor":true},{"citing_arxiv_id":"2604.27720","citing_title":"Auditing Frontier Vision-Language Models for Trustworthy Medical VQA: Grounding Failures, Format Collapse, and Domain Adaptation","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2605.09302","citing_title":"Discrete Langevin-Inspired Posterior Sampling","ref_index":44,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10218","citing_title":"Relative Score Policy Optimization for Diffusion Language Models","ref_index":97,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08144","citing_title":"NoiseRater: Meta-Learned Noise Valuation for Diffusion Model Training","ref_index":52,"is_internal_anchor":true},{"citing_arxiv_id":"2604.24763","citing_title":"Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation","ref_index":48,"is_internal_anchor":true},{"citing_arxiv_id":"2604.22152","citing_title":"dWorldEval: Scalable Robotic Policy Evaluation via Discrete Diffusion World Model","ref_index":45,"is_internal_anchor":true},{"citing_arxiv_id":"2604.10784","citing_title":"TorchUMM: A Unified Multimodal Model Codebase for Evaluation, Analysis, and Post-training","ref_index":29,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/A5YWT5NPNYXLFFO6RCZBDF3ZZW","json":"https://pith.science/pith/A5YWT5NPNYXLFFO6RCZBDF3ZZW.json","graph_json":"https://pith.science/api/pith-number/A5YWT5NPNYXLFFO6RCZBDF3ZZW/graph.json","events_json":"https://pith.science/api/pith-number/A5YWT5NPNYXLFFO6RCZBDF3ZZW/events.json","paper":"https://pith.science/paper/A5YWT5NP"},"agent_actions":{"view_html":"https://pith.science/pith/A5YWT5NPNYXLFFO6RCZBDF3ZZW","download_json":"https://pith.science/pith/A5YWT5NPNYXLFFO6RCZBDF3ZZW.json","view_paper":"https://pith.science/paper/A5YWT5NP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.15809&json=true","fetch_graph":"https://pith.science/api/pith-number/A5YWT5NPNYXLFFO6RCZBDF3ZZW/graph.json","fetch_events":"https://pith.science/api/pith-number/A5YWT5NPNYXLFFO6RCZBDF3ZZW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/A5YWT5NPNYXLFFO6RCZBDF3ZZW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/A5YWT5NPNYXLFFO6RCZBDF3ZZW/action/storage_attestation","attest_author":"https://pith.science/pith/A5YWT5NPNYXLFFO6RCZBDF3ZZW/action/author_attestation","sign_citation":"https://pith.science/pith/A5YWT5NPNYXLFFO6RCZBDF3ZZW/action/citation_signature","submit_replication":"https://pith.science/pith/A5YWT5NPNYXLFFO6RCZBDF3ZZW/action/replication_record"}},"created_at":"2026-05-17T23:38:52.311957+00:00","updated_at":"2026-05-17T23:38:52.311957+00:00"}