[论文] HyperTool: Beyond Step-Wise Tool Calls for Tool-Augmented Agents

论文概要

研究领域: NLP
作者: Yaxin Du, Yifan Zhou, Yujie Ge, Jiajun Wang, Xianghe Pang, Shuo Tang, Tuney Zheng, Bryan Dai, Jian Yang, Siheng Chen
发布时间: 2026-06-11
arXiv: 2606.13663

中文摘要

工具增强的LLM智能体通常依赖逐步原子工具调用，造成执行粒度不匹配。我们引入HyperTool，一种统一的MCP风格可执行工具接口，改变模型可见的工具执行单元。模型通过代码块调用HyperTool，可调用现有工具、操作返回值并本地传递中间结果。在MCP-Universe上，HyperTool将Qwen3-32B的平均准确率从15.69%提升至35.29%，将Qwen3-8B从9.93%提升至33.33%。

原文摘要

Tool-augmented LLM agents commonly rely on step-wise atomic tool calls, creating an execution-granularity mismatch. We introduce HyperTool, a unified executable MCP-style tool interface that changes the model-visible unit of tool execution. A model invokes HyperTool with a code block that can call existing tools, manipulate returned values, and pass intermediate results locally. On MCP-Universe, HyperTool improves average accuracy from 15.69% to 35.29% on Qwen3-32B and from 9.93% to 33.33% on Qwen3-8B.

自动采集于 2026-06-14

#论文 #arXiv #NLP #小凯