Towards Concise and Adaptive Thinking in Large Reasoning Models: A Survey

TL;DR

This survey reviews recent advances in making large reasoning models more concise and adaptive, aiming to improve efficiency by reducing unnecessary reasoning steps and balancing fast and slow thinking based on input difficulty.

Contribution

It provides a comprehensive overview of methodologies, benchmarks, and challenges related to concise and adaptive reasoning in large models, highlighting future research directions.

Findings

Recent progress in concise reasoning techniques

Development of benchmarks for adaptive reasoning

Identification of challenges in practical deployment

Abstract

Large reasoning models (LRMs) like OpenAI o1 and DeepSeek R1 have demonstrated impressive performance on complex reasoning tasks like mathematics and programming with long Chain-of-Thought (CoT) reasoning sequences (slow-thinking), compared with traditional large language models (fast-thinking). However, these reasoning models also face a huge challenge that generating unnecessarily lengthy and redundant reasoning chains even for trivial questions. This phenomenon leads to a significant waste of inference resources, increases the response time for simple queries, and hinders the practical application of LRMs in real-world products. To this end, it is crucial to shorten lengthy reasoning chains and learn adaptive reasoning between fast and slow thinking based on input difficulty. In this survey, we provide a comprehensive overview of recent progress in concise and adaptive thinking for efficient reasoning of LRMs, including methodologies, benchmarks, and challenges for future exploration. We hope this survey can help researchers quickly understand the landscape of this field and inspire novel adaptive thinking ideas to facilitate better usage of LRMs.