paper-with-me

홈 › Papers

LEAD: Length-Efficient Adaptive and Dynamic Reasoning for Large Language Models

2026-05-10 · Songtao Wei, Yi Li, Zhikai Li, Xu Hu, Yuede Ji, Guanpeng Li, Feng Chen, Carl Yang, Zhichun Guo, Bingzhe Li arxiv

Large reasoning models, such as OpenAI o1 and DeepSeek-R1, tend to become increasingly verbose as their reasoning capabilities improve. These inflated Chain-of-Thought (CoT) trajectories often exceed what the underlying problems require, wasting compute, latency, and context budgets. While introducing length-based efficiency rewards during reinforcement learning offers a natural remedy, existing methods struggle with two fundamental challenges: the optimal balance between correctness and efficiency is non-stationary throughout training, and intrinsic reasoning budgets vary drastically across problems. Relying on static reward weights and global length constraints inevitably forces a compromise between degraded accuracy and unrealized compression. To overcome these limitations, we propose LEAD (Length-Efficient Adaptive and Dynamic reasoning), a method that replaces static heuristics with online, self-adaptive mechanisms. LEAD dynamically calibrates the correctness-efficiency trade-off at each step using a Potential-Scaled Instability, directing optimization capacity to the most informative learning signal. Furthermore, it estimates an adaptive per-problem target length online based on the model's own correct rollouts, applying a symmetric efficiency reward that penalizes both overthinking and over-compression. Evaluated on five mathematical reasoning benchmarks, LEAD achieves the highest accuracy and Accuracy-Efficiency Score among RL-trained efficient-reasoning methods while producing substantially shorter outputs than the base model.

📄 PDF Abstract BibTeX arXiv:2605.09806

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement LearningMathematical Reasoning

Similar Papers 제목 키워드 기반

Thinking Fast and Right: Balancing Accuracy and Reasoning Length with Adaptive Rewards

2025-05-23 · Jinyan Su, Claire Cardie

Large language models (LLMs) have demonstrated strong reasoning abilities in mathematical tasks, often enhanced through reinforcement learning (RL). However, RL-trained models frequently produce unnecessarily long reason…

Reinforcement Learning (RL)

Leash: Adaptive Length Penalty and Reward Shaping for Efficient Large Reasoning Model

2025-12-25 · Yanhao Li, Lu Ma, Jiaran Zhang, Lexiang Tang 외 arxiv

Existing approaches typically rely on fixed length penalties, but such penalties are hard to tune and fail to adapt to the evolving reasoning abilities of LLMs, leading to suboptimal trade-offs between accuracy and conci…

Reinforcement LearningMathematical ReasoningInstruction Following

Think in Blocks: Adaptive Reasoning from Direct Response to Deep Reasoning

2025-08-21 · Yekun Zhu, Guang Chen, Chengjun Mao arxiv

Large Language Models (LLMs) with chains-of-thought have demonstrated strong performance on an increasing range of tasks, particularly those involving complex logical reasoning. However, excessively long chains can lead …

Reinforcement LearningLogical Reasoning

AALC: Large Language Model Efficient Reasoning via Adaptive Accuracy-Length Control

2025-06-25 · Ruosen Li, Ziming Luo, Quan Zhang, Ruochen Li 외

Large reasoning models (LRMs) achieve impressive reasoning capabilities by generating lengthy chain-of-thoughts, but this "overthinking" incurs high latency and cost without commensurate accuracy gains. In this work, we …

Language ModelingLanguage ModellingLarge Language ModelMath

ARES: Multimodal Adaptive Reasoning via Difficulty-Aware Token-Level Entropy Shaping

2025-10-09 · Shuang Chen, Yue Guo, Yimeng Ye, Shijue Huang 외 arxiv

Recent advances in multimodal large reasoning models (MLRMs) have substantially improved their ability to solve complex textual and visual tasks. However, these models tend to overthink on simple problems, producing unne…