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

Hugging Face Daily Papers Papers
reasoning chain-of-thought efficiency reinforcement-learning language-models adaptive-reasoning

Summary

LEAD dynamically adapts reasoning efficiency during training by using online calibration of correctness-efficiency trade-offs and adaptive problem-specific length targets, improving mathematical reasoning accuracy and reducing output length.

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.
Original Article
View Cached Full Text

Cached at: 05/15/26, 12:21 AM

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

Source: https://huggingface.co/papers/2605.09806

Abstract

LEAD is a method that dynamically adapts reasoning efficiency during training by using online calibration of correctness-efficiency trade-offs and adaptive problem-specific length targets to improve mathematical reasoning accuracy and efficiency.

Largereasoning models, such as OpenAI o1 and DeepSeek-R1, tend to become increasingly verbose as their reasoning capabilities improve. These inflatedChain-of-Thought(CoT) trajectories often exceed what the underlying problems require, wasting compute, latency, and context budgets. While introducinglength-based efficiency rewardsduringreinforcement learningoffers 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 aPotential-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 symmetricefficiency rewardthat penalizes both overthinking and over-compression. Evaluated on fivemathematical reasoning benchmarks, LEAD achieves the highest accuracy andAccuracy-Efficiency Scoreamong RL-trained efficient-reasoning methods while producing substantially shorter outputs than the base model.

View arXiv pageView PDFGitHub2Add to collection

Get this paper in your agent:

hf papers read 2605\.09806

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.09806 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.09806 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.09806 in a Space README.md to link it from this page.

Collections including this paper1

Similar Articles

Enhanced and Efficient Reasoning in Large Learning Models

arXiv cs.AI

This paper proposes a method for improving reasoning in large language models by recoding data to explicitly represent relationships, enabling efficient principled reasoning with polynomial-time learnability for relational rules, which addresses hallucinations and supports sound reasoning across multiple calls.

Can RL Teach Long-Horizon Reasoning to LLMs? Expressiveness Is Key

Hugging Face Daily Papers

This paper introduces ScaleLogic, a framework demonstrating that RL training compute scales as a power law with reasoning depth in LLMs. It highlights that logical expressiveness is key to improving downstream transfer and training efficiency.