paper-with-me

홈 › Papers

ConMax: Confidence-Maximizing Compression for Efficient Chain-of-Thought Reasoning

2026-01-08 · Minda Hu, Zexuan Qiu, Zenan Xu, Kun Li, Bo Zhou, Irwin King arxiv

Recent breakthroughs in Large Reasoning Models (LRMs) have demonstrated that extensive Chain-of-Thought (CoT) generation is critical for enabling intricate cognitive behaviors, such as self-verification and backtracking, to solve complex tasks. However, this capability often leads to ``overthinking'', where models generate redundant reasoning paths that inflate computational costs without improving accuracy. While Supervised Fine-Tuning (SFT) on reasoning traces is a standard paradigm for the 'cold start' phase, applying existing compression techniques to these traces often compromises logical coherence or incurs prohibitive sampling costs. In this paper, we introduce ConMax (Confidence-Maximizing Compression), a novel reinforcement learning framework designed to automatically compress reasoning traces while preserving essential reasoning patterns. ConMax formulates compression as a reward-driven optimization problem, training a policy to prune redundancy by maximizing a weighted combination of answer confidence for predictive fidelity and thinking confidence for reasoning validity through a frozen auxiliary LRM. Extensive experiments across five reasoning datasets demonstrate that ConMax achieves a superior efficiency-performance trade-off. Specifically, it reduces inference length by 43% over strong baselines at the cost of a mere 0.7% dip in accuracy, proving its effectiveness in generating high-quality, efficient training data for LRMs.

📄 PDF Abstract BibTeX arXiv:2601.04973

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

ConCISE: Confidence-guided Compression in Step-by-step Efficient Reasoning

2025-05-08 · Ziqing Qiao, Yongheng Deng, Jiali Zeng, Dong Wang 외

Large Reasoning Models (LRMs) perform strongly in complex reasoning tasks via Chain-of-Thought (CoT) prompting, but often suffer from verbose outputs caused by redundant content, increasing computational overhead, and de…

From Long to Lean: Performance-aware and Adaptive Chain-of-Thought Compression via Multi-round Refinement

2025-09-26 · Jianzhi Yan, Le Liu, Youcheng Pan, Shiwei Chen 외 arxiv

Chain-of-Thought (CoT) reasoning improves performance on complex tasks but introduces significant inference latency due to verbosity. We propose Multiround Adaptive Chain-of-Thought Compression (MACC), a framework that l…

R-KV: Redundancy-aware KV Cache Compression for Training-Free Reasoning Models Acceleration

2025-05-30 · Zefan Cai, Wen Xiao, Hanshi Sun, Cheng Luo 외

Reasoning models have demonstrated impressive performance in self-reflection and chain-of-thought reasoning. However, they often produce excessively long outputs, leading to prohibitively large key-value (KV) caches duri…

Mathematical Reasoning

HybridThinker: Efficient Chain-of-Thought Reasoning via Compressed Memory and Transient Thought Steps

2026-06-02 · Xin Liu, Runsong Zhao, Xinyu Liu, Junhao Ruan 외 arxiv

Extended chain-of-thought (CoT) traces improve LLM reasoning but incur substantial computational and memory costs. While existing CoT compression methods mitigate this by condensing thought steps into compact representat…

Reasoning Models Can be Accurately Pruned Via Chain-of-Thought Reconstruction

2025-09-15 · Ryan Lucas, Kayhan Behdin, Zhipeng Wang, Qingquan Song 외 arxiv

Reasoning language models such as DeepSeek-R1 produce long chain-of-thought traces during inference time which make them costly to deploy at scale. We show that using compression techniques such as neural network pruning…

Network Pruning