打造高质量数学推理数据集，助力大模型突破复杂问题求解
DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoning
- 构建高难度数学题库，覆盖5-9级挑战题，确保题目真实且无重复
- 经多轮去污染验证，答案可追溯，支持强化学习奖励机制
- 模型训练后不仅数学能力顶尖，还能迁移到生物物理等跨学科领域
强化学习结合大语言模型在复杂推理中展现潜力，但受限于缺乏大规模、高难度、无污染且可验证的训练数据。为此，我们推出DeepMath-103K，一个涵盖广泛数学主题的大规模数据集，主要包含5-9级高难度题目，经过多基准测试严格去污染，并提供可验证的答案以支持基于规则的强化学习奖励机制。该数据集包含三种不同R1解法，适配监督微调（SFT）等多种训练范式。在该数据集上训练的模型，在多个高难度数学基准上达到领先水平，并展现出向生物学、物理学和化学等领域的泛化能力，证明其广泛有效性。数据已公开：https://huggingface.co/datasets/zwhe99/DeepMath-103K。
原文摘要 · Abstract (English)
Reinforcement learning (RL) with large language models shows promise in complex reasoning. However, its progress is hindered by the lack of large-scale training data that is sufficiently challenging, contamination-free and verifiable. To this end, we introduce DeepMath-103K, a large-scale mathematical dataset designed with high difficulty (primarily levels 5-9), rigorous decontamination against numerous benchmarks, and verifiable answers for rule-based RL reward. It further includes three distinct R1 solutions adaptable for diverse training paradigms such as supervised fine-tuning (SFT). Spanning a wide range of mathematical topics, DeepMath-103K fosters the development of generalizable and advancing reasoning. Notably, models trained on DeepMath-103K achieve state-of-the-art results on challenging mathematical benchmarks and demonstrate generalization beyond math such as biology, physics and chemistry, underscoring its broad efficacy. Data: https://huggingface.co/datasets/zwhe99/DeepMath-103K.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。