P1: Mastering Physics Olympiads with Reinforcement Learning
Jiacheng Chen, Qianjia Cheng, Fangchen Yu, Haiyuan Wan, Yuchen Zhang, Shenghe Zheng, Junchi Yao, Qingyang Zhang, Haonan He, Yun Luo, Yufeng Zhao, Futing Wang, Li Sheng, Chengxing Xie, Yuxin Zuo, Yizhuo Li, Wenxauan Zeng, Yulun Wu, Rui Huang, Dongzhan Zhou, Kai Chen, Yu Qiao
2025-11-18
Summary
This paper demonstrates a significant leap in the ability of artificial intelligence, specifically large language models, to perform complex physics reasoning, going beyond simply solving textbook problems to tackling challenges that require understanding how the real world works.
What's the problem?
Current large language models are good at pattern recognition and can solve problems if they've seen similar ones before, but they often struggle with true scientific reasoning – applying fundamental principles to new situations. Physics, because it's so closely tied to the physical world, is a particularly difficult test for these models. The challenge is to create an AI that can actually *understand* physics, not just memorize solutions.
What's the solution?
The researchers developed a new family of open-source language models called P1, specifically trained using a technique called reinforcement learning. This means the models learned by trying to solve physics problems and getting feedback on their answers. The most powerful model, P1-235B-A22B, achieved gold-medal level performance on the International Physics Olympiad, even surpassing human competitors in many cases. They also created a system called PhysicsMinions that works with the model to further improve its performance. These models also showed strong reasoning skills in math and coding, suggesting they aren't just good at physics, but at general problem-solving.
Why it matters?
This work is important because it shows that AI is getting closer to being a useful tool for scientific discovery. If AI can truly reason about physics, it could help researchers develop new technologies, solve complex problems in fields like engineering and materials science, and even make new breakthroughs in our understanding of the universe. The fact that these models are open-source also means other researchers can build upon this work and accelerate progress in the field.
Abstract
Recent progress in large language models (LLMs) has moved the frontier from puzzle-solving to science-grade reasoning-the kind needed to tackle problems whose answers must stand against nature, not merely fit a rubric. Physics is the sharpest test of this shift, which binds symbols to reality in a fundamental way, serving as the cornerstone of most modern technologies. In this work, we manage to advance physics research by developing large language models with exceptional physics reasoning capabilities, especially excel at solving Olympiad-level physics problems. We introduce P1, a family of open-source physics reasoning models trained entirely through reinforcement learning (RL). Among them, P1-235B-A22B is the first open-source model with Gold-medal performance at the latest International Physics Olympiad (IPhO 2025), and wins 12 gold medals out of 13 international/regional physics competitions in 2024/2025. P1-30B-A3B also surpasses almost all other open-source models on IPhO 2025, getting a silver medal. Further equipped with an agentic framework PhysicsMinions, P1-235B-A22B+PhysicsMinions achieves overall No.1 on IPhO 2025, and obtains the highest average score over the 13 physics competitions. Besides physics, P1 models also present great performance on other reasoning tasks like math and coding, showing the great generalibility of P1 series.