paper-with-me

홈 › Papers

Hierarchical Budget Policy Optimization for Adaptive Reasoning

2025-07-21 · Shangke Lyu, Linjuan Wu, Yuchen Yan, Xingyu Wu, Hao Li, Yongliang Shen, Peisheng Jiang, Weiming Lu, Jun Xiao, Yueting Zhuang arxiv

Large reasoning models achieve remarkable performance through extensive chain-of-thought generation, yet they suffer from a critical inefficiency: applying uniformly extensive reasoning regardless of problem complexity. We present Hierarchical Budget Policy Optimization (HBPO), a reinforcement learning framework that enables models to learn problem-specific reasoning depths without sacrificing capability. Unlike existing approaches that impose rigid constraints or rely on discrete mode selection, HBPO partitions the exploration space into budget-constrained hierarchies (512-2560 tokens), each with differentiated reward structures that preserve both efficiency incentives and reasoning capabilities. This design addresses a fundamental challenge in efficient reasoning training: traditional length penalties systematically bias models away from necessary long reasoning paths, causing exploration space collapse. Through hierarchical sampling and budget-aware rewards, HBPO maintains exploration diversity while teaching models to recognize when extended deliberation is warranted. Extensive experiments demonstrate that HBPO reduces average token usage by up to 60.6% while improving accuracy by 3.14% across four reasoning benchmarks. Most notably, HBPO exhibits emergent adaptive behavior where models automatically adjust reasoning depth based on problem complexity. Our results suggest that reasoning efficiency and capability are not inherently conflicting, and can be simultaneously optimized through appropriately structured hierarchical training that preserves exploration diversity.

📄 PDF Abstract BibTeX arXiv:2507.15844

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

Not All Turns Are Equally Hard: Adaptive Thinking Budgets For Efficient Multi-Turn Reasoning

2026-04-06 · Neharika Jali, Anupam Nayak, Gauri Joshi arxiv

As LLM reasoning performance plateau, improving inference-time compute efficiency is crucial to mitigate overthinking and long thinking traces even for simple queries. Prior approaches including length regularization, ad…

Mathematical Reasoning

Learning from the Irrecoverable: Error-Localized Policy Optimization for Tool-Integrated LLM Reasoning

2026-02-10 · Qiao Liang, Yuke Zhu, Chao Ge, Lei Yang 외 arxiv

Tool-integrated reasoning (TIR) enables LLM agents to solve tasks through planning, tool use, and iterative revision, but outcome-only reinforcement learning in this setting suffers from sparse, delayed rewards and weak …

Reinforcement Learning

The Easy, the Hard, and the Learnable: Confidence and Difficulty-Adaptive Policy Optimization for LLM Reasoning

2026-06-06 · Zhanke Zhou, Xiangyu Lu, Chentao Cao, Brando Miranda 외 arxiv

RL with verifiable rewards can substantially improve LLM reasoning, yet standard GRPO-style training often treats easy, hard, and learnable questions alike through uniform sampling and weighting, leading to inefficient c…

AdaSR: Adaptive Streaming Reasoning with Hierarchical Relative Policy Optimization

2026-06-12 · Junlong Tong, Wenqi Xu, Yingqi Fan, Anhao Zhao 외 arxiv

Large reasoning models typically follow a read-then-think paradigm: they observe the complete input, reason over a static context, and then produce the answer. Yet many real-world scenarios are inherently dynamic, such a…

Computational Efficiency

HTPO: Towards Exploration-Exploitation Balanced Policy Optimization via Hierarchical Token-level Objective Control

2026-05-08 · Xincheng Yao, Ruoqi Li, Cheng Chen, Daoxin Zhang 외 arxiv

Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a pivotal technique for enhancing the reasoning capabilities of Large Language Models (LLMs). However, the de facto practice of mainstream RL algorithm…

Reinforcement Learning