Cache-of-Thought: Master-Apprentice Framework for Cost-Effective Vision Language Model Reasoning
This addresses the problem of cost-effective reasoning for users of VLMs, offering a significant performance boost for smaller models, though it is incremental as it builds on existing VLM architectures.
The paper tackles the trade-off between response quality and cost in Vision Language Models (VLMs) by proposing Cache of Thought (CoT), a master-apprentice framework that uses cached results from large VLMs to aid smaller VLMs, increasing overall reasoning performance by up to 7.7% under the same budget and boosting apprentice VLM performance by up to 36.6%.
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 widely recognized 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 performance of apprentice VLMs by up to 36.6%. Our code is available at https://github.com/UIUC-MONET/Cache-of-Thoughts