arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2407.21778cs.AIcs.RO
收藏

Tulip Agent——使基于LLM的智能体能够利用大型工具库解决任务

Tulip Agent -- Enabling LLM-Based Agents to Solve Tasks Using Large Tool Libraries

Honda Research Institute Europe(本田欧洲研究院)

Felix Ocker, Daniel Tanneberg, Julian Eggert, Michael Gienger

AI总结:

Tulip Agent通过递归搜索可扩展向量存储工具库而非在系统提示中编码全部工具描述,使LLM智能体能够高效使用大型工具库,降低推理成本并支持工具集自适应扩展。

摘要:

我们介绍了Tulip Agent,这是一种面向自主LLM智能体的架构,该智能体对可能包含大量工具的工具库具有创建、读取、更新和删除(Create, Read, Update, and Delete)访问权限。与当前最先进的实现不同,Tulip Agent不会在系统提示中编码所有可用工具的描述,因为这会占用模型的上下文窗口,也不会嵌入整个提示以检索合适的工具。相反,Tulip Agent可以在其可扩展的工具库中递归搜索合适的工具,该工具库示例性地实现为向量存储(vector store)。Tulip Agent架构显著降低了推理成本,允许使用甚至大型工具库,并使智能体能够适应和扩展其工具集。我们在数学背景下通过若干消融研究评估了该架构,并通过对机器人学的应用展示了其通用性。参考实现和基准可在github.com/HRI-EU/tulip_agent获取。

英文摘要:

We introduce tulip agent, an architecture for autonomous LLM-based agents with Create, Read, Update, and Delete access to a tool library containing a potentially large number of tools. In contrast to state-of-the-art implementations, tulip agent does not encode the descriptions of all available tools in the system prompt, which counts against the model's context window, or embed the entire prompt for retrieving suitable tools. Instead, the tulip agent can recursively search for suitable tools in its extensible tool library, implemented exemplarily as a vector store. The tulip agent architecture significantly reduces inference costs, allows using even large tool libraries, and enables the agent to adapt and extend its set of tools. We evaluate the architecture with several ablation studies in a mathematics context and demonstrate its generalizability with an application to robotics. A reference implementation and the benchmark are available at github.com/HRI-EU/tulip_agent.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑