Cache-of-Thought: Master-Apprentice Framework for Cost-Effective Vision Language Model Reasoning

Vision Language Models (VLMs) have achieved remarkable success in a wide range of vision applications of increasing complexity and scales, yet choosing the right VLM model size involves a trade-off between response quality and cost. While smaller VLMs are cheaper to run, they typically produce responses only marginally better than random guessing on benchmarks such as MMMU. In this paper, we propose Cache of Thought (CoT), a master-apprentice framework for collaborative inference between large and small VLMs. CoT manages high-quality query results from large VLMs (master) in a cache, which are then selected via a novel multi-modal retrieval and in-context learning to aid the performance of small VLMs (apprentice). We extensively evaluate CoT on various widelyrecognized and challenging general reasoning benchmarks, and show that CoT increases overall reasoning performance by up to 7.7% under the same budget, and specifically boosts the reasoning performance of apprentice VLMs by up to 36.6%. Our code is available at https://github.com/UIUC-MONET/Cache -of-Thoughts . Stage 1 Master VLM Search In-context Example Now you should answer the following question given the above example: Top-k Retrieve Stage 2 Apprentice VLM Groups of men from three different areas of the country are to be tested for mean weight. The entries in Table 12 .13 are the weights for the different groups. <image 1> What is the Mean Square Factor? A.