arXiv cs.AI· Xinyan Wang, Xiaogeng Liu, Ming Pei, Chaowei Xiao·· 14 小时前AI 评分45
ROM：通过流式检测与干预实时缓解大推理模型过度思考
ROM: Real-time Overthinking Mitigation via Streaming Detection and Intervention
AI 导读
研究者提出 ROM（Real-time Overthinking Mitigation），用仅占骨干参数 0.07–0.25% 的轻量流式检测器监控冻结的大推理模型，在推理边界实时干预以缓解过度思考。
正文 · 原文
Abstract:Large reasoning models (LRMs) often suffer from overthinking, continuing with redundant verification and repeated attempts even after reaching a correct solution. This wastes computation and can overturn an initially correct answer. We show that the transition from productive to redundant reasoning is decodable from late-layer hidden states. These representations separate the two phases around first-correct-solution boundaries beyond what position and lexical controls explain, and distinguish correct from incorrect solution endings. Building on this finding, we propose ROM (Real-time Overthinking Mitigation), which uses a lightweight streaming detector (0.07--0.25\% of backbone parameters) to monitor a frozen LRM and intervene at well-formed reasoning boundaries without answer extraction or probe decoding. Counterfactual Self-Correction (CSC) augments detector training with wrong$\rightarrow$correct trajectories to preserve useful self-correction. Heads are trained separately for each backbone using shared MATH500 supervision and transfer zero-shot to four unseen benchmarks. Evaluated against ten baselines across five backbones from three families and five benchmarks, ROM$_{\text{CSC}}$ reduces response length by 28--77\% (45\% on average), with significant reductions in all 25 settings and accuracy matching or exceeding vanilla decoding in 22. Measurements on two backbones further show 20--70\% lower end-to-end latency with less than 0.5\% detector overhead per decoding step. Code is available at this https URL.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2603.22016 [cs.LG] |
| (or arXiv:2603.22016v4 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2603.22016 arXiv-issued DOI via DataCite |
Submission history
From: Xinyan Wang [view email]
[v1]
Mon, 23 Mar 2026 14:26:57 UTC (6,373 KB)
[v2]
Sat, 9 May 2026 22:19:27 UTC (936 KB)
[v3]
Mon, 10 Aug 2026 05:45:00 UTC (343 KB)
[v4]
Thu, 29 Oct 2026 01:29:42 UTC (770 KB)
来源：arXiv cs.AI · arxiv.org