Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs
This addresses efficiency issues for users of large language models by mitigating overthinking, though it is incremental as it builds on existing self-training paradigms.
The paper tackles the problem of overthinking in o1-like LLMs, where excessive computational resources are used for simple problems with minimal benefit, and shows that their approach reduces computational overhead while preserving accuracy across benchmarks like GSM8K and MATH500.
The remarkable performance of models like the OpenAI o1 can be attributed to their ability to emulate human-like long-time thinking during inference. These models employ extended chain-of-thought (CoT) processes, exploring multiple strategies to enhance problem-solving capabilities. However, a critical question remains: How to intelligently and efficiently scale computational resources during testing. This paper presents the first comprehensive study on the prevalent issue of overthinking in these models, where excessive computational resources are allocated for simple problems with minimal benefit. We introduce novel efficiency metrics from both outcome and process perspectives to evaluate the rational use of computational resources by o1-like models. Using a self-training paradigm, we propose strategies to mitigate overthinking, streamlining reasoning processes without compromising accuracy. Experimental results show that our approach successfully reduces computational overhead while preserving model performance across a range of testsets with varying difficulty levels, such as GSM8K, MATH500, GPQA, and AIME.