Chain of Images for Intuitively Reasoning
This addresses the problem of enhancing logical reasoning in AI models for researchers and developers by enabling multimodal reasoning, though it appears incremental as it builds on existing Chain of Thoughts methods with a visual extension.
The paper tackles the limitation of current Large Language Models (LLMs) in using visual intuition for reasoning by proposing a Chain of Images (CoI) approach that converts complex language problems into simple pattern recognition through generated image sequences, and experiments on Geometry, Chess, and Common Sense tasks show significant performance improvements over pure-language Chain of Thoughts baselines.
The human brain is naturally equipped to comprehend and interpret visual information rapidly. When confronted with complex problems or concepts, we use flowcharts, sketches, and diagrams to aid our thought process. Leveraging this inherent ability can significantly enhance logical reasoning. However, current Large Language Models (LLMs) do not utilize such visual intuition to help their thinking. Even the most advanced version language models (e.g., GPT-4V and LLaVA) merely align images into textual space, which means their reasoning processes remain purely verbal. To mitigate such limitations, we present a Chain of Images (CoI) approach, which can convert complex language reasoning problems to simple pattern recognition by generating a series of images as intermediate representations. Furthermore, we have developed a CoI evaluation dataset encompassing 15 distinct domains where images can intuitively aid problem-solving. Based on this dataset, we aim to construct a benchmark to assess the capability of future multimodal large-scale models to leverage images for reasoning. In supporting our CoI reasoning, we introduce a symbolic multimodal large language model (SyMLLM) that generates images strictly based on language instructions and accepts both text and image as input. Experiments on Geometry, Chess and Common Sense tasks sourced from the CoI evaluation dataset show that CoI improves performance significantly over the pure-language Chain of Thoughts (CoT) baselines. The code is available at https://github.com/GraphPKU/CoI.