BMMR: A Large-Scale Bilingual Multimodal Multi-Discipline Reasoning Dataset

TL;DR

BMMR is a comprehensive large-scale bilingual multimodal dataset designed to evaluate and improve large multimodal models' reasoning across multiple disciplines, with extensive experiments revealing current limitations and potential improvements.

Contribution

The paper introduces BMMR, a novel large-scale bilingual multimodal reasoning dataset with high-quality reasoning paths, and proposes a process-based verifier for detailed evaluation.

Findings

State-of-the-art models still have significant room for improvement on BMMR-Eval.

Models show discipline bias, performing better on certain subjects.

Fine-tuning on BMMR-Train reduces performance gaps between models.

Abstract

In this paper, we introduce BMMR, a large-scale bilingual, multimodal, multi-disciplinary reasoning dataset for the community to develop and evaluate large multimodal models (LMMs). BMMR comprises 110k college-level questions spanning 300 UNESCO-defined subjects, spanning diverse formats-multiple-choice, fill-in-the-blank, and open-ended QA-and sourced from both print and digital media such as books, exams, and quizzes. All data are curated and filtered via a human-in-the-loop and scalable framework, and each instance is paired with a high-quality reasoning path. The dataset is organized into two parts: BMMR-Eval that comprises 20,458 high-quality instances to comprehensively assess LMMs' knowledge and reasoning across multiple disciplines in both Chinese and English; and BMMR-Train that contains 88,991 instances to support further research and development, extending the current focus on mathematical reasoning to diverse disciplines and domains. In addition, we propose the process-based multi-discipline verifier (i.e., BMMR-Verifier) for accurate and fine-grained evaluation of reasoning paths. Extensive experiments on 24 models reveal that (i) even SOTA models (e.g., o3 and Gemini-2.5-Pro) leave substantial headroom on BMMR-Eval; (ii) reasoning models exhibit discipline bias and outperform LMMs only on specific subjects; (iii) open-source models still trail their proprietary counterparts; and (iv) fine-tuning on BMMR-Train narrows this gap. Additionally, we conduct reasoning-chain analyses using BMMR-Verifier and other in-depth studies, uncovering the challenges LMMs currently face in multidisciplinary reasoning. We will release the data, and we hope our work can offer insights and contributions to the community.