HDFlow：通过混合思维与动态工作流增强 LLM 复杂问题求解能力
HDFlow: Enhancing LLM Complex Problem-Solving with Hybrid Thinking and Dynamic Workflows
- Tencent AI Lab（腾讯AI实验室）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
针对LLM复杂推理受限的问题，提出HDFlow框架，通过动态工作流分解任务并结合快慢混合思维，显著提升了模型的复杂问题求解能力与计算效率。
AI中文摘要：
尽管大型语言模型（LLMs）近期取得了进展，但其在需要多步思维和结合各种技能的复杂推理问题上的表现仍然有限。为了解决这一问题，我们提出了一种用于 LLM 复杂推理的新框架 HDFlow，它以自适应方式结合了快慢两种思维模式。我们的方法包含两个关键组件：1）一种名为 Dynamic Workflow 的慢速深思熟虑推理新方法，它自动将复杂问题分解为更易处理的子任务，并动态设计工作流以组装专门的 LLM 或符号推理工具来解决子任务；2） Hybrid Thinking，一个根据问题复杂度动态结合快慢思维的一般框架。最后，我们提出了一种易于扩展的方法，用于自动合成包含 27K 道挑战性推理问题的大规模数据集以用于复杂推理，并提出了一种混合思维微调方法，在该数据集上训练较小的 LLMs 以内化快/慢混合推理策略。在四个推理基准数据集上的实验表明，我们带有动态工作流的慢思维显著优于 Chain-of-Thought，而混合思维在提供计算效率与性能的有效平衡的同时实现了最高准确率。使用我们的混合思维方法进行微调也显著提升了开源语言模型的复杂推理能力。结果展示了慢思维、动态工作流和混合思维在扩展 LLM 复杂问题求解前沿方面的潜力。
英文摘要：
Despite recent advancements in large language models (LLMs), their performance on complex reasoning problems requiring multi-step thinking and combining various skills is still limited. To address this, we propose a novel framework HDFlow for complex reasoning with LLMs that combines fast and slow thinking modes in an adaptive manner. Our approach consists of two key components: 1) a new approach for slow, deliberate reasoning called Dynamic Workflow, which automatically decomposes complex problems into more manageable sub-tasks and dynamically designs a workflow to assemble specialized LLM or symbolic reasoning tools to solve sub-tasks; 2) Hybrid Thinking, a general framework that dynamically combines fast and slow thinking based on problem complexity. Finally, we propose an easy-to-scale method for automatically synthesizing a large-scale dataset of 27K challenging reasoning problems for complex reasoning and a hybrid thinking tuning method that trains smaller LLMs on this dataset to internalize the fast/slow hybrid reasoning strategies. Experiments on four reasoning benchmark datasets demonstrate that our slow thinking with dynamic workflows significantly outperforms Chain-of-Thought, and hybrid thinking achieves the highest accuracy while providing an effective balance between computational efficiency and performance. Fine-tuning using our hybrid thinking approach also significantly boosts the complex reasoning capabilities of open-source language models. The results showcase the promise of slow thinking, dynamic workflows, and hybrid thinking in expanding the frontier of complex problem-solving with LLMs\footnote{Code and data will be released at \url{https://github.com/wenlinyao/HDFlow}.}.
