STAR: STacked AutoRegressive Scheme for Unified Multimodal Learning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866912767070437376 |
|---|---|
| author | Qin, Jie Huang, Jiancheng Qiao, Limeng Ma, Lin |
| author_facet | Qin, Jie Huang, Jiancheng Qiao, Limeng Ma, Lin |
| contents | Multimodal large language models (MLLMs) play a pivotal role in advancing the quest for general artificial intelligence. However, achieving unified target for multimodal understanding and generation remains challenging due to optimization conflicts and performance trade-offs. To effectively enhance generative performance while preserving existing comprehension capabilities, we introduce STAR: a STacked AutoRegressive scheme for task-progressive unified multimodal learning. This approach decomposes multimodal learning into multiple stages: understanding, generation, and editing. By freezing the parameters of the fundamental autoregressive (AR) model and progressively stacking isomorphic AR modules, it avoids cross-task interference while expanding the model's capabilities. Concurrently, we introduce a high-capacity VQ to enhance the granularity of image representations and employ an implicit reasoning mechanism to improve generation quality under complex conditions. Experiments demonstrate that STAR achieves state-of-the-art performance on GenEval (0.91), DPG-Bench (87.44), and ImgEdit (4.34), validating its efficacy for unified multimodal learning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_13752 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | STAR: STacked AutoRegressive Scheme for Unified Multimodal Learning Qin, Jie Huang, Jiancheng Qiao, Limeng Ma, Lin Computer Vision and Pattern Recognition Artificial Intelligence Multimodal large language models (MLLMs) play a pivotal role in advancing the quest for general artificial intelligence. However, achieving unified target for multimodal understanding and generation remains challenging due to optimization conflicts and performance trade-offs. To effectively enhance generative performance while preserving existing comprehension capabilities, we introduce STAR: a STacked AutoRegressive scheme for task-progressive unified multimodal learning. This approach decomposes multimodal learning into multiple stages: understanding, generation, and editing. By freezing the parameters of the fundamental autoregressive (AR) model and progressively stacking isomorphic AR modules, it avoids cross-task interference while expanding the model's capabilities. Concurrently, we introduce a high-capacity VQ to enhance the granularity of image representations and employ an implicit reasoning mechanism to improve generation quality under complex conditions. Experiments demonstrate that STAR achieves state-of-the-art performance on GenEval (0.91), DPG-Bench (87.44), and ImgEdit (4.34), validating its efficacy for unified multimodal learning. |
| title | STAR: STacked AutoRegressive Scheme for Unified Multimodal Learning |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2512.13752 |