---
id: 20260925-T0-05
title: "让推理模型少说废话：冗余感知学习压缩推理链长度"
title_en: "Redundancy-Aware Learning Cuts Unnecessarily Long Reasoning Traces"
url: https://ai.daily.yangsir.net/daily/20260925-T0-05
issue_date: 2026-09-25
publish_date: 2026-09-24T04:00:00.000Z
category: research
source_name: "arXiv cs.CL (NLP)"
source_url: https://arxiv.org/abs/2609.27156
---

# 让推理模型少说废话：冗余感知学习压缩推理链长度

大型推理模型常生成正确但过长的推理链。现有方法通过轨迹级目标或局部的token和步骤级信号提升效率，但很少建模步骤之间的语义依赖。该研究提出冗余感知学习方法，针对步骤间的语义冗余进行优化。效果是推理链更短，同时保持答案正确性。对实际部署而言，这意味着更低的推理成本和更快的响应速度。

## English Version

**Redundancy-Aware Learning Cuts Unnecessarily Long Reasoning Traces**

Large reasoning models often produce correct but unnecessarily long reasoning traces. Existing methods improve efficiency via trajectory-level objectives or local token- and step-level signals, but rarely model inter-step semantic dependencies. This work proposes a redundancy-aware learning approach that targets semantic redundancy between steps. The result is shorter reasoning traces while preserving answer correctness, which translates to lower inference cost and faster responses in deployment.

---

**来源**：[arXiv cs.CL (NLP)](https://arxiv.org/abs/2609.27156)

**详情页**：https://ai.daily.yangsir.net/daily/20260925-T0-05

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*