Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2602.09443 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914318730133504 |
|---|---|
| author | Luo, Yun Wang, Futing Cheng, Qianjia Yu, Fangchen Lei, Haodi Yan, Jianhao Li, Chenxi Chen, Jiacheng Zhao, Yufeng Wan, Haiyuan Zhang, Yuchen Zheng, Shenghe Yao, Junchi Zhang, Qingyang He, Haonan Zeng, Wenxuan Sheng, Li Xie, Chengxing Zuo, Yuxin Li, Yizhuo Wu, Yulun Huang, Rui Zhou, Dongzhan Chen, Kai Qiao, Yu Bai, Lei Cheng, Yu Ding, Ning Zhou, Bowen Ye, Peng Cui, Ganqu |
| author_facet | Luo, Yun Wang, Futing Cheng, Qianjia Yu, Fangchen Lei, Haodi Yan, Jianhao Li, Chenxi Chen, Jiacheng Zhao, Yufeng Wan, Haiyuan Zhang, Yuchen Zheng, Shenghe Yao, Junchi Zhang, Qingyang He, Haonan Zeng, Wenxuan Sheng, Li Xie, Chengxing Zuo, Yuxin Li, Yizhuo Wu, Yulun Huang, Rui Zhou, Dongzhan Chen, Kai Qiao, Yu Bai, Lei Cheng, Yu Ding, Ning Zhou, Bowen Ye, Peng Cui, Ganqu |
| contents | The transition from symbolic manipulation to science-grade reasoning represents a pivotal frontier for Large Language Models (LLMs), with physics serving as the critical test anchor for binding abstract logic to physical reality. Physics demands that a model maintain physical consistency with the laws governing the universe, a task that fundamentally requires multimodal perception to ground abstract logic in reality. At the Olympiad level, diagrams are often constitutive rather than illustrative, containing essential constraints, such as boundary conditions and spatial symmetries, that are absent from the text. To bridge this visual-logical gap, we introduce P1-VL, a family of open-source vision-language models engineered for advanced scientific reasoning. Our method harmonizes Curriculum Reinforcement Learning, which employs progressive difficulty expansion to stabilize post-training, with Agentic Augmentation, enabling iterative self-verification at inference. Evaluated on HiPhO, a rigorous benchmark of 13 exams from 2024-2025, our flagship P1-VL-235B-A22B becomes the first open-source Vision-Language Model (VLM) to secure 12 gold medals and achieves the state-of-the-art performance in the open-source models. Our agent-augmented system achieves the No.2 overall rank globally, trailing only Gemini-3-Pro. Beyond physics, P1-VL demonstrates remarkable scientific reasoning capacity and generalizability, establishing significant leads over base models in STEM benchmarks. By open-sourcing P1-VL, we provide a foundational step toward general-purpose physical intelligence to better align visual perceptions with abstract physical laws for machine scientific discovery. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_09443 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | P1-VL: Bridging Visual Perception and Scientific Reasoning in Physics Olympiads Luo, Yun Wang, Futing Cheng, Qianjia Yu, Fangchen Lei, Haodi Yan, Jianhao Li, Chenxi Chen, Jiacheng Zhao, Yufeng Wan, Haiyuan Zhang, Yuchen Zheng, Shenghe Yao, Junchi Zhang, Qingyang He, Haonan Zeng, Wenxuan Sheng, Li Xie, Chengxing Zuo, Yuxin Li, Yizhuo Wu, Yulun Huang, Rui Zhou, Dongzhan Chen, Kai Qiao, Yu Bai, Lei Cheng, Yu Ding, Ning Zhou, Bowen Ye, Peng Cui, Ganqu Artificial Intelligence The transition from symbolic manipulation to science-grade reasoning represents a pivotal frontier for Large Language Models (LLMs), with physics serving as the critical test anchor for binding abstract logic to physical reality. Physics demands that a model maintain physical consistency with the laws governing the universe, a task that fundamentally requires multimodal perception to ground abstract logic in reality. At the Olympiad level, diagrams are often constitutive rather than illustrative, containing essential constraints, such as boundary conditions and spatial symmetries, that are absent from the text. To bridge this visual-logical gap, we introduce P1-VL, a family of open-source vision-language models engineered for advanced scientific reasoning. Our method harmonizes Curriculum Reinforcement Learning, which employs progressive difficulty expansion to stabilize post-training, with Agentic Augmentation, enabling iterative self-verification at inference. Evaluated on HiPhO, a rigorous benchmark of 13 exams from 2024-2025, our flagship P1-VL-235B-A22B becomes the first open-source Vision-Language Model (VLM) to secure 12 gold medals and achieves the state-of-the-art performance in the open-source models. Our agent-augmented system achieves the No.2 overall rank globally, trailing only Gemini-3-Pro. Beyond physics, P1-VL demonstrates remarkable scientific reasoning capacity and generalizability, establishing significant leads over base models in STEM benchmarks. By open-sourcing P1-VL, we provide a foundational step toward general-purpose physical intelligence to better align visual perceptions with abstract physical laws for machine scientific discovery. |
| title | P1-VL: Bridging Visual Perception and Scientific Reasoning in Physics Olympiads |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2602.09443 |