EvoLMM: Self-Evolving Large Multimodal Models with Continuous Rewards
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_ | 1866917337104384000 |
|---|---|
| author | Thawakar, Omkar Venkatraman, Shravan Thawkar, Ritesh Shaker, Abdelrahman Cholakkal, Hisham Anwer, Rao Muhammad Khan, Salman Khan, Fahad |
| author_facet | Thawakar, Omkar Venkatraman, Shravan Thawkar, Ritesh Shaker, Abdelrahman Cholakkal, Hisham Anwer, Rao Muhammad Khan, Salman Khan, Fahad |
| contents | Recent advances in large multimodal models (LMMs) have enabled impressive reasoning and perception abilities, yet most existing training pipelines still depend on human-curated data or externally verified reward models, limiting their autonomy and scalability. In this work, we strive to improve LMM reasoning capabilities in a purely unsupervised fashion (without any annotated data or reward distillation). To this end, we propose a self-evolving framework, named EvoLMM, that instantiates two cooperative agents from a single backbone model: a Proposer, which generates diverse, image-grounded questions, and a Solver, which solves them through internal consistency, where learning proceeds through a continuous self-rewarding process. This dynamic feedback encourages both the generation of informative queries and the refinement of structured reasoning without relying on ground-truth or human judgments. When using the popular Qwen2.5-VL as the base model, our EvoLMM yields consistent gains upto $\sim$3\% on multimodal math-reasoning benchmarks, including ChartQA, MathVista, and MathVision, using only raw training images. We hope our simple yet effective approach will serve as a solid baseline easing future research in self-improving LMMs in a fully-unsupervised fashion. Our code and models are available at https://github.com/mbzuai-oryx/EvoLMM. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_16672 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | EvoLMM: Self-Evolving Large Multimodal Models with Continuous Rewards Thawakar, Omkar Venkatraman, Shravan Thawkar, Ritesh Shaker, Abdelrahman Cholakkal, Hisham Anwer, Rao Muhammad Khan, Salman Khan, Fahad Computer Vision and Pattern Recognition Recent advances in large multimodal models (LMMs) have enabled impressive reasoning and perception abilities, yet most existing training pipelines still depend on human-curated data or externally verified reward models, limiting their autonomy and scalability. In this work, we strive to improve LMM reasoning capabilities in a purely unsupervised fashion (without any annotated data or reward distillation). To this end, we propose a self-evolving framework, named EvoLMM, that instantiates two cooperative agents from a single backbone model: a Proposer, which generates diverse, image-grounded questions, and a Solver, which solves them through internal consistency, where learning proceeds through a continuous self-rewarding process. This dynamic feedback encourages both the generation of informative queries and the refinement of structured reasoning without relying on ground-truth or human judgments. When using the popular Qwen2.5-VL as the base model, our EvoLMM yields consistent gains upto $\sim$3\% on multimodal math-reasoning benchmarks, including ChartQA, MathVista, and MathVision, using only raw training images. We hope our simple yet effective approach will serve as a solid baseline easing future research in self-improving LMMs in a fully-unsupervised fashion. Our code and models are available at https://github.com/mbzuai-oryx/EvoLMM. |
| title | EvoLMM: Self-Evolving Large Multimodal Models with Continuous Rewards |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2511.16672 |