P1-VL: Bridging Visual Perception and Scientific Reasoning in Physics Olympiads

2026-02-10

AI Summary

This work addresses the challenge in physics olympiad problems where critical constraints are embedded in diagrams while textual descriptions are absent, leading to a disconnect between visual perception and physical reasoning. To bridge this gap, the authors propose P1-VL, an open-source vision-language model that integrates curriculum-based reinforcement learning—featuring progressively increasing problem difficulty—with an agent-driven self-verification mechanism at inference time. This enables end-to-end multimodal perception to high-level physical reasoning. Evaluated on the HiPhO benchmark, P1-VL achieves state-of-the-art performance among open-source vision-language models, securing 12 gold medals and ranking second globally, just behind Gemini-3-Pro, while also demonstrating strong generalization across diverse STEM domains.

Technology Category

• Computer Vision: Low Level & Physics-based Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

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.

Problem

• scientific reasoning
• visual perception
• physics Olympiads
• multimodal reasoning
• physical consistency

Innovation

• Vision-Language Model
• Scientific Reasoning
• Curriculum Reinforcement Learning
• Agentic Augmentation
• Physical Intelligence