MathVista

MathVista is a benchmark dataset used to evaluate the reasoning ability of multi-modal large language models (MLLMs) without requiring ground truth labels.

Papers using MathVista (26)

Qwen3-VL Technical Report2025 · 12 citesInternvl3.5: Advancing Open-source Multimodal Models In Versatility, Reasoning, And Efficiency2025 · 2 citesVision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models2025 · 1 citesTVI-CoT: Text-Visual Interleaved Chain-of-Thought Reasoning for Multimodal Understanding2026Difference Feedback: Generating Multimodal Process-Level Supervision for VLM Reinforcement Learning2026Credit Where It is Due: Cross-Modality Connectivity Drives Precise Reinforcement Learning for MLLM Reasoning2026EvoLMM: Self-Evolving Large Multimodal Models with Continuous Rewards2025ChainV: Atomic Visual Hints Make Multimodal Reasoning Shorter and Better2025Diagnosing Visual Reasoning: Challenges, Insights, and a Path Forward2025VOLD: Reasoning Transfer from LLMs to Vision-Language Models via On-Policy Distillation2025SAIL-VL2 Technical Report2025Training Vision-Language Process Reward Models for Test-Time Scaling in Multimodal Reasoning: Key Insights and Lessons Learned2025Qianfan-vl: Domain-enhanced Universal Vision-language Models2025CoRGI: Verified Chain-of-Thought Reasoning with Post-hoc Visual Grounding2025Athena: Enhancing Multimodal Reasoning with Data-efficient Process Reward Models2025Mmjee-eval: A Bilingual Multimodal Benchmark For Evaluating Scientific Reasoning In Vision-language Models2025SRPO: Enhancing Multimodal LLM Reasoning via Reflection-Aware Reinforcement Learning2025Advancing Multimodal Reasoning via Reinforcement Learning with Cold Start2025First SFT, Second RL, Third UPT: Continual Improving Multi-Modal LLM Reasoning via Unsupervised Post-Training2025Skywork R1V: Pioneering Multimodal Reasoning with Chain-of-Thought2025OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL Cycles2025Evolutionary Prompt Optimization Discovers Emergent Multimodal Reasoning Strategies in Vision-Language Models20252.5 Years in Class: A Multimodal Textbook for Vision-Language Pretraining2025MathScape: Benchmarking Multimodal Large Language Models in Real-World Mathematical Contexts2024CMM-Math: A Chinese Multimodal Math Dataset To Evaluate and Enhance the Mathematics Reasoning of Large Multimodal Models2024 · 1 citesText as Images: Can Multimodal Large Language Models Follow Printed Instructions in Pixels?2023