{"work":{"id":"1393dc24-a6b2-44e1-b5d7-7009d1fa4811","openalex_id":"https://openalex.org/W4403662644","doi":"10.48550/arxiv.2408.12528","arxiv_id":"2408.12528","raw_key":null,"title":"Show-o: One Single Transformer to Unify Multimodal Understanding and Generation","authors":null,"authors_text":"Jinheng Xie, Weijia Mao, Zechen Bai, David Junhao Zhang, Weihao Wang, Kevin Qinghong Lin","year":2024,"venue":"cs.CV","abstract":"We present a unified transformer, i.e., Show-o, that unifies multimodal understanding and generation. Unlike fully autoregressive models, Show-o unifies autoregressive and (discrete) diffusion modeling to adaptively handle inputs and outputs of various and mixed modalities. The unified model flexibly supports a wide range of vision-language tasks including visual question-answering, text-to-image generation, text-guided inpainting/extrapolation, and mixed-modality generation. Across various benchmarks, it demonstrates comparable or superior performance to existing individual models with an equivalent or larger number of parameters tailored for understanding or generation. This significantly highlights its potential as a next-generation foundation model. Code and models are released at https://github.com/showlab/Show-o.","external_url":"https://arxiv.org/abs/2408.12528","cited_by_count":3,"metadata_source":"pith","metadata_fetched_at":"2026-08-05T02:28:24.338817+00:00","pith_arxiv_id":"2408.12528","created_at":"2026-05-09T06:25:47.548589+00:00","updated_at":"2026-08-05T02:28:24.338817+00:00","title_quality_ok":true,"display_title":"Show-o: One Single Transformer to Unify Multimodal Understanding and Generation","render_title":"Show-o: One Single Transformer to Unify Multimodal Understanding and Generation"},"hub":{"state":{"work_id":"1393dc24-a6b2-44e1-b5d7-7009d1fa4811","tier":"super_hub","tier_reason":"100+ Pith inbound or 10,000+ external citations","pith_inbound_count":120,"external_cited_by_count":3,"distinct_field_count":9,"first_pith_cited_at":"2024-04-22T17:56:09+00:00","last_pith_cited_at":"2026-07-01T03:11:21+00:00","author_build_status":"needed","summary_status":"needed","contexts_status":"needed","graph_status":"needed","ask_index_status":"needed","reader_status":"not_needed","recognition_status":"not_needed","updated_at":"2026-08-23T21:19:23.999995+00:00","tier_text":"super_hub"},"tier":"super_hub","role_counts":[{"context_role":"background","n":17},{"context_role":"baseline","n":13},{"context_role":"method","n":3},{"context_role":"dataset","n":1}],"polarity_counts":[{"context_polarity":"background","n":17},{"context_polarity":"baseline","n":13},{"context_polarity":"use_method","n":3},{"context_polarity":"use_dataset","n":1}],"runs":{"ask_index":{"job_type":"ask_index","status":"succeeded","result":{"title":"Show-o: One Single Transformer to Unify Multimodal Understanding and Generation","claims":[{"claim_text":"We present a unified transformer, i.e., Show-o, that unifies multimodal understanding and generation. Unlike fully autoregressive models, Show-o unifies autoregressive and (discrete) diffusion modeling to adaptively handle inputs and outputs of various and mixed modalities. The unified model flexibly supports a wide range of vision-language tasks including visual question-answering, text-to-image generation, text-guided inpainting/extrapolation, and mixed-modality generation. Across various benchmarks, it demonstrates comparable or superior performance to existing individual models with an equ","claim_type":"abstract","evidence_strength":"source_metadata"},{"claim_text":"categories: single object, two objects, counting, colors, position, and attribute binding. 28 Nucleus-ImageTECHNICALREPORT Figure 11: GenEval overall scores for top-performing models. Nucleus-Image matches Qwen-Image at 0.87 and leads all models on spatial position understanding. Table 9: GenEval results. Model Single Two Count Colors Position Attr. Overall Show-o [44] 0.95 0.52 0.49 0.82 0.11 0.28 0.53 PixArt-α[45] 0.98 0.50 0.44 0.80 0.08 0.07 0.48 Emu3-Gen [46] 0.98 0.71 0.34 0.81 0.17 0.21 0","claim_type":"baseline","confidence":0.95,"evidence_strength":"citation_context"},{"claim_text":"IMU-to-4D (ours), red:MobilePoser. Ours aligns more closely with the GT. vocabulary with special tokens that mark the start and end of each IMU token, as well as the start and end of each output modality (text, motion, and scene). 3.3 Architecture Our architecture is illustrated in Fig. 2. We use Phi-1.5 as our backbone and use pretrained checkpoints from Show-o [93]. We adopt the 1.5B-parameter variant to balance efficiency and performance. Multimodal Integration.To handle multiple input modali","claim_type":"method","confidence":0.95,"evidence_strength":"citation_context"},{"claim_text":"Unified MetaQuery-XL [88]✓ ✓ ✓ ✓ ✓ SEED-X [32]✓ ✓ ✓ ✓ ✓ TokenFlow-XL [92]✓ ✓ ✓ ✓ ILLUME [110]✓ ✓ ✓ ✓ ✓ InternVL-U [103]✓ ✓ ✓ ✓ ✓ UniVideo [119]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓✓ Native Unified Chameleon [100]✓ ✓ ✓ ✓ L WM [69]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ Janus [123]✓ ✓ ✓ ✓ Janus-Pro [14]✓ ✓ ✓ ✓ Transfusion [149]✓ ✓ ✓ ✓ Emu3 [115]✓ ✓ ✓△ △ △✓ ✓ Show-o [133]✓ ✓ ✓ ✓ ✓ Show-o2 [134]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓△ Bagel [22]✓ ✓ ✓ ✓ ✓ ✓✓ Mogao [63]✓ ✓ ✓ ✓△ △ HaploOmni [132]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ VILA-U [130]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ HunyuanImage 3.0 [8]△ ","claim_type":"baseline","confidence":0.9,"evidence_strength":"citation_context"},{"claim_text":"0093.3397.1787.7883.21 83.58 GPT Image 1 [64]89.1588.2990.75 89.66 91.3387.08 84.57 84.5796.32 97.3288.55 88.35 87.07 89.44 87.2283.9685.59 83.2190.0093.3389.83 86.8389.7393.46 Qwen-Image [106]86.14 86.8386.18 87.22 90.5091.5088.2290.7879.81 79.3879.30 80.88 79.21 78.94 78.85 81.69 75.57 78.59100.00 100.0092.76 89.1490.3091.42 AR based Models Show-o [113]59.72 58.8673.08 75.83 74.83 79.83 78.82 78.32 65.57 69.3253.67 50.38 60.95 56.82 68.59 68.96 66.46 56.22 63.33 66.67 3.83 2.8355.02 50.92 Infi","claim_type":"baseline","confidence":0.9,"evidence_strength":"citation_context"},{"claim_text":"InternVL [15] scaled up vision foundation models for alignment with LLMs. Additionally, GPT-4V [54] and Gemini [68] demonstrated strong gen- eral visual understanding. Mixture-of-Experts (MoE) ap- proaches improved understanding in DeepSeek-VL2 [81], Uni-MoE [34], MoV A [97], and MoME [63]. Some mod- els, including SEED-X [20], Chameleon [67], Show-o [83], Transfusion [95], and Janus [14, 80], unified vision under- standing and generation. However, most existing LMMs still lack reasoning capabil","claim_type":"background","confidence":0.9,"evidence_strength":"citation_context"},{"claim_text":"(a) LLaDA-V demonstrates superior performance on more bench- marks compared to LLaMA3-V when trained on the same dataset, particularly excelling in multi- disciplinary knowledge and mathematical reasoning tasks. (b) LLaDA-V achieves state-of-the-art performance in multimodal understanding among both hybrid autoregressive-diffusion (such as MetaMorph [31] and Show-o [28]) and purely diffusion-based models. modeling, its capabilities and potential in multimodal understanding remain largely unexplo","claim_type":"baseline","confidence":0.9,"evidence_strength":"citation_context"}],"why_cited":"Pith tracks Show-o: One Single Transformer to Unify Multimodal Understanding and Generation because it crossed a citation-hub threshold. Current citing contexts most often use it as background evidence (17 contexts).","role_counts":[{"n":17,"context_role":"background"},{"n":13,"context_role":"baseline"},{"n":3,"context_role":"method"},{"n":1,"context_role":"dataset"}]},"error":null,"updated_at":"2026-07-01T10:11:16.886530+00:00"},"author_expand":{"job_type":"author_expand","status":"succeeded","result":{"authors_linked":[{"id":"9bfa6f75-273f-481c-9584-96a65427609a","orcid":null,"display_name":"Jinheng Xie"},{"id":"347c7274-bc93-4dc7-9bb4-978010bb403f","orcid":null,"display_name":"Weijia Mao"},{"id":"7a5d822a-5f68-4cc0-bf8b-cdcba4d96d84","orcid":null,"display_name":"Zechen Bai"},{"id":"2961e04b-d0e6-4584-8dc0-03834c91f906","orcid":null,"display_name":"David Junhao Zhang"},{"id":"9c99f165-65a7-4c40-ae83-f8aa00e51bb0","orcid":null,"display_name":"Weihao Wang"},{"id":"75a87808-ba8e-41dd-870b-ddc13d473e7a","orcid":null,"display_name":"Kevin Qinghong Lin"}]},"error":null,"updated_at":"2026-07-01T10:11:16.881417+00:00"},"context_extract":{"job_type":"context_extract","status":"succeeded","result":{"enqueued_papers":25},"error":null,"updated_at":"2026-05-14T17:48:53.527237+00:00"},"graph_features":{"job_type":"graph_features","status":"succeeded","result":{"co_cited":[{"title":"Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling","work_id":"67d9e391-26d1-459e-ab56-07e60511c886","shared_citers":27},{"title":"Emerging Properties in Unified Multimodal Pretraining","work_id":"e0cfd82c-f5d4-44fd-b531-ec73ab0a805b","shared_citers":21},{"title":"Emu3: Next-Token Prediction is All You Need","work_id":"720d288e-fac0-464c-9929-19efd9a52afc","shared_citers":20},{"title":"Chameleon: Mixed-Modal Early-Fusion Foundation Models","work_id":"2661b9a6-25cc-41a1-8100-612d2b801289","shared_citers":15},{"title":"SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis","work_id":"8034c587-fba6-4941-87ba-c98f2ac962cb","shared_citers":14},{"title":"BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset","work_id":"86d896d2-592f-4d9b-938e-dfeb11f9388f","shared_citers":13},{"title":"Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model","work_id":"c2bb4d2d-29de-4bf2-9150-3d6373ff358f","shared_citers":13},{"title":"DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models","work_id":"c5006563-f3ec-438a-9e35-b7b484f34828","shared_citers":12},{"title":"GPT-4 Technical Report","work_id":"b928e041-6991-4c08-8c81-0359e4097c7b","shared_citers":11},{"title":"OmniGen2: Towards Instruction-Aligned Multimodal Generation","work_id":"d3153e5f-b6e2-4ab3-9f41-e24e24d64496","shared_citers":11},{"title":"Qwen2.5-VL Technical Report","work_id":"69dffacb-bfe8-442d-be86-48624c60426f","shared_citers":11},{"title":"ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment","work_id":"94248955-4bc5-4517-98a0-66224a36d865","shared_citers":10},{"title":"Show-o2: Improved Native Unified Multimodal Models","work_id":"77f00563-1ce6-4fba-9d4e-c8ce83f716ac","shared_citers":10},{"title":"Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution","work_id":"8abcfe4f-e0fb-44b7-9123-448fac95f90a","shared_citers":9},{"title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","work_id":"e6b75ad5-2877-4168-97c8-710407094d20","shared_citers":8},{"title":"Gemini: A Family of Highly Capable Multimodal Models","work_id":"83f7c85b-3f11-450f-ac0c-64d9745220b2","shared_citers":8},{"title":"Qwen-Image Technical Report","work_id":"d06d7ecc-7579-4f89-a60b-4278a0f3c562","shared_citers":8},{"title":"Seed-x: Multimodal models with unified multi-granularity comprehension and generation","work_id":"15953092-dd9e-49ae-9f72-e28fc93a6068","shared_citers":8},{"title":"UniWorld-V1: High-Resolution Semantic Encoders for Unified Visual Understanding and Generation","work_id":"488a273e-95d8-46f1-87c7-2244068d00d0","shared_citers":8},{"title":"Wu et al","work_id":"8aa59998-5ac6-4ee1-bcb4-cf7fe479058b","shared_citers":8},{"title":"Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation","work_id":"41efe203-9377-4c63-b1d6-e499cd6e46f6","shared_citers":7},{"title":"Flow Matching for Generative Modeling","work_id":"6edb71c4-5d64-40af-a394-9757ea051a36","shared_citers":7},{"title":"GPT-4o System Card","work_id":"f37bf1c7-4964-4e56-9762-d20da8d9009f","shared_citers":7},{"title":"Hierarchical Text-Conditional Image Generation with CLIP Latents","work_id":"0c6a768b-70b8-4242-bb0e-459f1008c9fc","shared_citers":7}],"time_series":[{"n":2,"year":2024},{"n":6,"year":2025},{"n":31,"year":2026}],"dependency_candidates":[]},"error":null,"updated_at":"2026-05-14T17:48:32.369670+00:00"},"identity_refresh":{"job_type":"identity_refresh","status":"succeeded","result":{"items":[{"title":"Qwen3 Technical Report","outcome":"unchanged","work_id":"25a4e30c-1232-48e7-9925-02fa12ba7c9e","resolver":"local_arxiv","confidence":0.98,"old_work_id":"25a4e30c-1232-48e7-9925-02fa12ba7c9e"}],"counts":{"fixed":0,"merged":0,"unchanged":1,"quarantined":0,"needs_external_resolution":0},"errors":[],"attempted":1},"error":null,"updated_at":"2026-05-14T17:48:36.430879+00:00"},"role_polarity":{"job_type":"role_polarity","status":"succeeded","result":{"title":"Show-o: One Single Transformer to Unify Multimodal Understanding and Generation","claims":[{"claim_text":"We present a unified transformer, i.e., Show-o, that unifies multimodal understanding and generation. Unlike fully autoregressive models, Show-o unifies autoregressive and (discrete) diffusion modeling to adaptively handle inputs and outputs of various and mixed modalities. The unified model flexibly supports a wide range of vision-language tasks including visual question-answering, text-to-image generation, text-guided inpainting/extrapolation, and mixed-modality generation. Across various benchmarks, it demonstrates comparable or superior performance to existing individual models with an equ","claim_type":"abstract","evidence_strength":"source_metadata"},{"claim_text":"categories: single object, two objects, counting, colors, position, and attribute binding. 28 Nucleus-ImageTECHNICALREPORT Figure 11: GenEval overall scores for top-performing models. Nucleus-Image matches Qwen-Image at 0.87 and leads all models on spatial position understanding. Table 9: GenEval results. Model Single Two Count Colors Position Attr. Overall Show-o [44] 0.95 0.52 0.49 0.82 0.11 0.28 0.53 PixArt-α[45] 0.98 0.50 0.44 0.80 0.08 0.07 0.48 Emu3-Gen [46] 0.98 0.71 0.34 0.81 0.17 0.21 0","claim_type":"baseline","confidence":0.95,"evidence_strength":"citation_context"},{"claim_text":"IMU-to-4D (ours), red:MobilePoser. Ours aligns more closely with the GT. vocabulary with special tokens that mark the start and end of each IMU token, as well as the start and end of each output modality (text, motion, and scene). 3.3 Architecture Our architecture is illustrated in Fig. 2. We use Phi-1.5 as our backbone and use pretrained checkpoints from Show-o [93]. We adopt the 1.5B-parameter variant to balance efficiency and performance. Multimodal Integration.To handle multiple input modali","claim_type":"method","confidence":0.95,"evidence_strength":"citation_context"},{"claim_text":"Unified MetaQuery-XL [88]✓ ✓ ✓ ✓ ✓ SEED-X [32]✓ ✓ ✓ ✓ ✓ TokenFlow-XL [92]✓ ✓ ✓ ✓ ILLUME [110]✓ ✓ ✓ ✓ ✓ InternVL-U [103]✓ ✓ ✓ ✓ ✓ UniVideo [119]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓✓ Native Unified Chameleon [100]✓ ✓ ✓ ✓ L WM [69]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ Janus [123]✓ ✓ ✓ ✓ Janus-Pro [14]✓ ✓ ✓ ✓ Transfusion [149]✓ ✓ ✓ ✓ Emu3 [115]✓ ✓ ✓△ △ △✓ ✓ Show-o [133]✓ ✓ ✓ ✓ ✓ Show-o2 [134]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓△ Bagel [22]✓ ✓ ✓ ✓ ✓ ✓✓ Mogao [63]✓ ✓ ✓ ✓△ △ HaploOmni [132]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ VILA-U [130]✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ HunyuanImage 3.0 [8]△ ","claim_type":"baseline","confidence":0.9,"evidence_strength":"citation_context"},{"claim_text":"0093.3397.1787.7883.21 83.58 GPT Image 1 [64]89.1588.2990.75 89.66 91.3387.08 84.57 84.5796.32 97.3288.55 88.35 87.07 89.44 87.2283.9685.59 83.2190.0093.3389.83 86.8389.7393.46 Qwen-Image [106]86.14 86.8386.18 87.22 90.5091.5088.2290.7879.81 79.3879.30 80.88 79.21 78.94 78.85 81.69 75.57 78.59100.00 100.0092.76 89.1490.3091.42 AR based Models Show-o [113]59.72 58.8673.08 75.83 74.83 79.83 78.82 78.32 65.57 69.3253.67 50.38 60.95 56.82 68.59 68.96 66.46 56.22 63.33 66.67 3.83 2.8355.02 50.92 Infi","claim_type":"baseline","confidence":0.9,"evidence_strength":"citation_context"},{"claim_text":"InternVL [15] scaled up vision foundation models for alignment with LLMs. Additionally, GPT-4V [54] and Gemini [68] demonstrated strong gen- eral visual understanding. Mixture-of-Experts (MoE) ap- proaches improved understanding in DeepSeek-VL2 [81], Uni-MoE [34], MoV A [97], and MoME [63]. Some mod- els, including SEED-X [20], Chameleon [67], Show-o [83], Transfusion [95], and Janus [14, 80], unified vision under- standing and generation. However, most existing LMMs still lack reasoning capabil","claim_type":"background","confidence":0.9,"evidence_strength":"citation_context"},{"claim_text":"(a) LLaDA-V demonstrates superior performance on more bench- marks compared to LLaMA3-V when trained on the same dataset, particularly excelling in multi- disciplinary knowledge and mathematical reasoning tasks. (b) LLaDA-V achieves state-of-the-art performance in multimodal understanding among both hybrid autoregressive-diffusion (such as MetaMorph [31] and Show-o [28]) and purely diffusion-based models. modeling, its capabilities and potential in multimodal understanding remain largely unexplo","claim_type":"baseline","confidence":0.9,"evidence_strength":"citation_context"}],"why_cited":"Pith tracks Show-o: One Single Transformer to Unify Multimodal Understanding and Generation because it crossed a citation-hub threshold. Current citing contexts most often use it as background evidence (17 contexts).","role_counts":[{"n":17,"context_role":"background"},{"n":13,"context_role":"baseline"},{"n":3,"context_role":"method"},{"n":1,"context_role":"dataset"}]},"error":null,"updated_at":"2026-07-01T10:11:16.884015+00:00"},"summary_claims":{"job_type":"summary_claims","status":"succeeded","result":{"title":"Show-o: One Single Transformer to Unify Multimodal Understanding and Generation","claims":[{"claim_text":"We present a unified transformer, i.e., Show-o, that unifies multimodal understanding and generation. Unlike fully autoregressive models, Show-o unifies autoregressive and (discrete) diffusion modeling to adaptively handle inputs and outputs of various and mixed modalities. The unified model flexibly supports a wide range of vision-language tasks including visual question-answering, text-to-image generation, text-guided inpainting/extrapolation, and mixed-modality generation. Across various benchmarks, it demonstrates comparable or superior performance to existing individual models with an equ","claim_type":"abstract","evidence_strength":"source_metadata"}],"why_cited":"Pith tracks Show-o: One Single Transformer to Unify Multimodal Understanding and Generation because it crossed a citation-hub threshold.","role_counts":[]},"error":null,"updated_at":"2026-05-14T17:48:53.543205+00:00"}},"summary":{"title":"Show-o: One Single Transformer to Unify Multimodal Understanding and Generation","claims":[{"claim_text":"We present a unified transformer, i.e., Show-o, that unifies multimodal understanding and generation. Unlike fully autoregressive models, Show-o unifies autoregressive and (discrete) diffusion modeling to adaptively handle inputs and outputs of various and mixed modalities. The unified model flexibly supports a wide range of vision-language tasks including visual question-answering, text-to-image generation, text-guided inpainting/extrapolation, and mixed-modality generation. Across various benchmarks, it demonstrates comparable or superior performance to existing individual models with an equ","claim_type":"abstract","evidence_strength":"source_metadata"}],"why_cited":"Pith tracks Show-o: One Single Transformer to Unify Multimodal Understanding and Generation because it crossed a citation-hub threshold.","role_counts":[]},"graph":{"co_cited":[{"title":"Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling","work_id":"67d9e391-26d1-459e-ab56-07e60511c886","shared_citers":27},{"title":"Emerging Properties in Unified Multimodal Pretraining","work_id":"e0cfd82c-f5d4-44fd-b531-ec73ab0a805b","shared_citers":21},{"title":"Emu3: Next-Token Prediction is All You Need","work_id":"720d288e-fac0-464c-9929-19efd9a52afc","shared_citers":20},{"title":"Chameleon: Mixed-Modal Early-Fusion Foundation Models","work_id":"2661b9a6-25cc-41a1-8100-612d2b801289","shared_citers":15},{"title":"SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis","work_id":"8034c587-fba6-4941-87ba-c98f2ac962cb","shared_citers":14},{"title":"BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset","work_id":"86d896d2-592f-4d9b-938e-dfeb11f9388f","shared_citers":13},{"title":"Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model","work_id":"c2bb4d2d-29de-4bf2-9150-3d6373ff358f","shared_citers":13},{"title":"DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models","work_id":"c5006563-f3ec-438a-9e35-b7b484f34828","shared_citers":12},{"title":"GPT-4 Technical Report","work_id":"b928e041-6991-4c08-8c81-0359e4097c7b","shared_citers":11},{"title":"OmniGen2: Towards Instruction-Aligned Multimodal Generation","work_id":"d3153e5f-b6e2-4ab3-9f41-e24e24d64496","shared_citers":11},{"title":"Qwen2.5-VL Technical Report","work_id":"69dffacb-bfe8-442d-be86-48624c60426f","shared_citers":11},{"title":"ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment","work_id":"94248955-4bc5-4517-98a0-66224a36d865","shared_citers":10},{"title":"Show-o2: Improved Native Unified Multimodal Models","work_id":"77f00563-1ce6-4fba-9d4e-c8ce83f716ac","shared_citers":10},{"title":"Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution","work_id":"8abcfe4f-e0fb-44b7-9123-448fac95f90a","shared_citers":9},{"title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","work_id":"e6b75ad5-2877-4168-97c8-710407094d20","shared_citers":8},{"title":"Gemini: A Family of Highly Capable Multimodal Models","work_id":"83f7c85b-3f11-450f-ac0c-64d9745220b2","shared_citers":8},{"title":"Qwen-Image Technical Report","work_id":"d06d7ecc-7579-4f89-a60b-4278a0f3c562","shared_citers":8},{"title":"Seed-x: Multimodal models with unified multi-granularity comprehension and generation","work_id":"15953092-dd9e-49ae-9f72-e28fc93a6068","shared_citers":8},{"title":"UniWorld-V1: High-Resolution Semantic Encoders for Unified Visual Understanding and Generation","work_id":"488a273e-95d8-46f1-87c7-2244068d00d0","shared_citers":8},{"title":"Wu et al","work_id":"8aa59998-5ac6-4ee1-bcb4-cf7fe479058b","shared_citers":8},{"title":"Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation","work_id":"41efe203-9377-4c63-b1d6-e499cd6e46f6","shared_citers":7},{"title":"Flow Matching for Generative Modeling","work_id":"6edb71c4-5d64-40af-a394-9757ea051a36","shared_citers":7},{"title":"GPT-4o System Card","work_id":"f37bf1c7-4964-4e56-9762-d20da8d9009f","shared_citers":7},{"title":"Hierarchical Text-Conditional Image Generation with CLIP Latents","work_id":"0c6a768b-70b8-4242-bb0e-459f1008c9fc","shared_citers":7}],"time_series":[{"n":2,"year":2024},{"n":6,"year":2025},{"n":31,"year":2026}],"dependency_candidates":[]},"authors":[{"id":"2961e04b-d0e6-4584-8dc0-03834c91f906","orcid":null,"display_name":"David Junhao Zhang","source":"manual","import_confidence":0.72},{"id":"9bfa6f75-273f-481c-9584-96a65427609a","orcid":null,"display_name":"Jinheng Xie","source":"manual","import_confidence":0.72},{"id":"75a87808-ba8e-41dd-870b-ddc13d473e7a","orcid":null,"display_name":"Kevin Qinghong Lin","source":"manual","import_confidence":0.72},{"id":"9c99f165-65a7-4c40-ae83-f8aa00e51bb0","orcid":null,"display_name":"Weihao Wang","source":"manual","import_confidence":0.72},{"id":"347c7274-bc93-4dc7-9bb4-978010bb403f","orcid":null,"display_name":"Weijia Mao","source":"manual","import_confidence":0.72},{"id":"7a5d822a-5f68-4cc0-bf8b-cdcba4d96d84","orcid":null,"display_name":"Zechen Bai","source":"manual","import_confidence":0.72}]}}