LogicVista: Multimodal LLM Logical Reasoning Benchmark in Visual Contexts
This work addresses the need for better assessment of logical reasoning in MLLMs, which is crucial for applications like navigation and puzzle-solving, but it is incremental as it primarily provides a new evaluation benchmark.
The authors tackled the lack of systematic evaluation for multimodal large language models' logical reasoning in visual contexts by introducing LogicVista, a benchmark with 448 multiple-choice questions across 5 tasks, and found that current models show limited proficiency in these areas.
We propose LogicVista, an evaluation benchmark that assesses the integrated logical reasoning capabilities of multimodal large language models (MLLMs) in Visual contexts. Recent advancements in MLLMs have demonstrated various fascinating abilities, from crafting poetry based on an image to performing mathematical reasoning. However, there is still a lack of systematic evaluation of MLLMs' proficiency in logical reasoning tasks, which are essential for activities like navigation and puzzle-solving. Thus we evaluate general logical cognition abilities across 5 logical reasoning tasks encompassing 9 different capabilities, using a sample of 448 multiple-choice questions. Each question is annotated with the correct answer and the human-written reasoning behind the selection, enabling both open-ended and multiple-choice evaluation. A total of 8 MLLMs are comprehensively evaluated using LogicVista. Code and Data Available at https://github.com/Yijia-Xiao/LogicVista.