H²R：用分层提示奖励改进数学推理强化学习
H$^2$R: Hierarchical Hint-Based Reward for Mathematical Reasoning via Reinforcement Learning
一篇 RL 训练新方法：把数学题拆成子问题当奖励里程碑，Qwen2.5-Math-7B 在 HiMath 上从 30.4% 涨到 45.6%
论文提出 Hierarchical Hint-based Reward（H²R），把数学题自带的子问题结构当作 RLVR 训练中的中间条件信号。配套发布了由分层结构数学题组成的 HiMath 基准。Qwen2.5-Math-7B 在 HiMath 上的准确率从 30.4% 提升到 39.1%，使用 Oracle Hierarchical Hints 时可达 45.6%。分层提示即使不做任务专项训练，也能提升多种 LLM 的推理表现。
H$^2$R: Hierarchical Hint-Based Reward for Mathematical Reasoning via Reinforcement Learning
Reinforcement Learning with Verifiable Rewards has shown great promise in enhancing mathematical reasoning in large language models. However, existing approaches provide limited intermediate supervision for complex, multi-step reasoning tasks. To bridge this gap, we propose \textbf{Hierarchical Hint-based Reward}, a framework that leverages the intrinsic sub-question structure of mathematical problems as structured intermediate conditioning signals during RLVR training. Unlike chain-of-thought prompting, these hierarchical hints serve as logically grounded intermediate milestones that effectively reduce the reasoning horizon toward the final answer. To validate this framework, we introduce \textbf{HiMath}, a benchmark of hierarchically structured mathematical problems. Across diverse LLMs, hierarchical hints consistently improve reasoning performance even without task-specific training. When integrated into RL training via H$^2$R, they boost Qwen2.5-Math-7B accuracy on HiMath from 30.4\% to 39.1\% using hierarchical hints, and up to 45.6\% with Oracle Hierarchical Hints. This pronounced improvement demonstrates that incorporating reliable intermediate checkpoints into reward design is essential for sustaining complex mathematical reasoning.