Harnessing the Reasoning Economy: A Survey of Efficient Reasoning for Large Language Models
Rui Wang, Hongru Wang, Boyang Xue, Jianhui Pang, Shudong Liu, Yi Chen, Jiahao Qiu, Derek Fai Wong, Heng Ji, Kam-Fai Wong
2025-04-02
Summary
This paper explores how to make AI models think more efficiently, saving computing power and time while still getting good results.
What's the problem?
AI models can be inefficient in their reasoning, using too much computing power and taking too long to solve problems.
What's the solution?
This paper surveys different ways to make AI reasoning more efficient, by analyzing why it's inefficient, studying different reasoning patterns, and suggesting potential solutions.
Why it matters?
This work matters because it can lead to faster and more practical AI models that can be used in real-world applications.
Abstract
Recent advancements in Large Language Models (LLMs) have significantly enhanced their ability to perform complex reasoning tasks, transitioning from fast and intuitive thinking (System 1) to slow and deep reasoning (System 2). While System 2 reasoning improves task accuracy, it often incurs substantial computational costs due to its slow thinking nature and inefficient or unnecessary reasoning behaviors. In contrast, System 1 reasoning is computationally efficient but leads to suboptimal performance. Consequently, it is critical to balance the trade-off between performance (benefits) and computational costs (budgets), giving rise to the concept of reasoning economy. In this survey, we provide a comprehensive analysis of reasoning economy in both the post-training and test-time inference stages of LLMs, encompassing i) the cause of reasoning inefficiency, ii) behavior analysis of different reasoning patterns, and iii) potential solutions to achieve reasoning economy. By offering actionable insights and highlighting open challenges, we aim to shed light on strategies for improving the reasoning economy of LLMs, thereby serving as a valuable resource for advancing research in this evolving area. We also provide a public repository to continually track developments in this fast-evolving field.