CLAIAug 30, 2024

MultiMath: Bridging Visual and Mathematical Reasoning for Large Language Models

arXiv:2409.00147v145 citationsh-index: 9Has Code
Originality Incremental advance
AI Analysis

This addresses the need for models that can handle visual mathematical tasks like diagrams and charts, which is incremental as it builds on existing multimodal and math reasoning research.

The authors tackled the problem of large language models lacking integration of visual inputs for mathematical reasoning by introducing MultiMath-7B, a multimodal model that achieves state-of-the-art performance on existing multimodal benchmarks and excels on text-only ones.

The rapid development of large language models (LLMs) has spurred extensive research into their domain-specific capabilities, particularly mathematical reasoning. However, most open-source LLMs focus solely on mathematical reasoning, neglecting the integration with visual injection, despite the fact that many mathematical tasks rely on visual inputs such as geometric diagrams, charts, and function plots. To fill this gap, we introduce \textbf{MultiMath-7B}, a multimodal large language model that bridges the gap between math and vision. \textbf{MultiMath-7B} is trained through a four-stage process, focusing on vision-language alignment, visual and math instruction-tuning, and process-supervised reinforcement learning. We also construct a novel, diverse and comprehensive multimodal mathematical dataset, \textbf{MultiMath-300K}, which spans K-12 levels with image captions and step-wise solutions. MultiMath-7B achieves state-of-the-art (SOTA) performance among open-source models on existing multimodal mathematical benchmarks and also excels on text-only mathematical benchmarks. Our model and dataset are available at {\textcolor{blue}{\url{https://github.com/pengshuai-rin/MultiMath}}}.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes