evothink-evolving-thinking-in-large-reasoning-mode

name	evothink-evolving-thinking-in-large-reasoning-mode
description	Skill generated from arXiv paper 2607.19962: EvoThink: Evolving Thinking in Large Reasoning Models via Self-Pruning and Aha-Moment Preference Optimization
metadata	{"arxiv":{"id":"2607.19962","title":"EvoThink: Evolving Thinking in Large Reasoning Models via Self-Pruning and Aha-Moment Preference Optimization","authors":["Xinbang Dai","Zheyu Xin","Huikang Hu","Lin Ren","Rihui Jin","Guohui Xiao","Guilin Qi","Kuicai Dong","Zhaocheng Du","Yuyang Zhang"],"published":"2026-07-22","categories":["cs.AI"],"url":"https://arxiv.org/abs/2607.19962","utility":1}}

EvoThink: Evolving Thinking in Large Reasoning Models via Self-Pruning and Aha-Moment Preference Optimization

arXiv: 2607.19962
Published: 2026-07-22
Authors: Xinbang Dai, Zheyu Xin, Huikang Hu, Lin Ren, Rihui Jin, Guohui Xiao, Guilin Qi, Kuicai Dong, Zhaocheng Du, Yuyang Zhang
Categories: cs.AI
Utility: 1.00

Key Innovation

Large Reasoning Models (LRMs) often suffer from overthinking due to redundant verification steps. Existing approaches for mitigating overthinking, such as fast-slow thinking switching and reasoning trajectory compression, fail to make a fine-grained distinction between beneficial and redundant steps within the LRM's reasoning process, and may thus impair reasoning capability in their pursuit of efficiency. To simultaneously improve reasoning efficiency and capability, we propose EvoThink, a fram...

Potential Application

This paper presents advancements that could be applied to enhance agent capabilities in the areas of cs.AI.

References

• arXiv: https://arxiv.org/abs/2607.19962