MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI
This benchmark addresses the problem of evaluating multimodal AI models on expert-level tasks for researchers and developers, though it is incremental as it builds on existing benchmarking efforts.
The authors introduced MMMU, a benchmark with 11.5K multimodal questions from college-level disciplines to evaluate models on advanced perception and reasoning, finding that top models like GPT-4V and Gemini Ultra achieved only 56% and 59% accuracy, indicating significant challenges.
We introduce MMMU: a new benchmark designed to evaluate multimodal models on massive multi-discipline tasks demanding college-level subject knowledge and deliberate reasoning. MMMU includes 11.5K meticulously collected multimodal questions from college exams, quizzes, and textbooks, covering six core disciplines: Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, and Tech & Engineering. These questions span 30 subjects and 183 subfields, comprising 30 highly heterogeneous image types, such as charts, diagrams, maps, tables, music sheets, and chemical structures. Unlike existing benchmarks, MMMU focuses on advanced perception and reasoning with domain-specific knowledge, challenging models to perform tasks akin to those faced by experts. The evaluation of 14 open-source LMMs as well as the proprietary GPT-4V(ision) and Gemini highlights the substantial challenges posed by MMMU. Even the advanced GPT-4V and Gemini Ultra only achieve accuracies of 56% and 59% respectively, indicating significant room for improvement. We believe MMMU will stimulate the community to build next-generation multimodal foundation models towards expert artificial general intelligence.