CATP-LLM: Empowering Large Language Models for Cost-Aware Tool Planning

ID: 6e6c027e
原文链接: https://arxiv.org/abs/2411.16313
PDF: https://arxiv.org/pdf/2411.16313
作者: Duo Wu, Jinghe Wang, Yuan Meng, Yanning Zhang, Le Sun, Zhi Wang
日期 / 版本: Submitted on 25 Nov 2024 (v1), last revised 13 Jul 2025 (this version, v3)
分类: agents
来源类型: paper
arXiv 分类: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
标签: llm-agent, tool-planning, cost-aware, tool-use, benchmark, 2411-16313
质量评分: 4/5
抓取时间: 2026-07-27T04:25:26.452521+00:00

中文导读

论文关注 LLM 工具规划中的执行成本问题摘要指出既有研究常忽略工具耗时等成本，可能生成收益低于代价的计划；CATP-LLM 框架首次系统引入 cost-aware tool planning，让 LLM 在调度外部工具时同时考虑任务效果和执行成本

为什么值得关注

CATP-LLM 把工具执行成本纳入规划目标，避免 Agent 生成性能收益抵不过成本的昂贵计划

这篇论文值得放进 AAIF 的原因，是它把 Agent 系统里的一个具体工程问题讲清楚：论文关注 LLM 工具规划中的执行成本问题摘要指出既有研究常忽略工具耗时等成本，可能生成收益低于代价的计划；CATP-LLM 框架首次系统引入 cost-aware tool planning，让 LLM 在调度外部工具时同时考虑任务效果和执行成本。以下内容基于 arXiv 页面元数据、摘要与条目已有摘要整理，未补充论文摘要之外的实验细节。

关键信息

论文标题：CATP-LLM: Empowering Large Language Models for Cost-Aware Tool Planning
作者：Duo Wu, Jinghe Wang, Yuan Meng, Yanning Zhang, Le Sun, Zhi Wang
arXiv：https://arxiv.org/abs/2411.16313
发布时间 / 修订：Submitted on 25 Nov 2024 (v1), last revised 13 Jul 2025 (this version, v3)
arXiv 分类：Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
arXiv 备注：Accepted to ICCV 2025. Codes and dataset are available from the arXiv-linked repository.
关联标签：llm-agent, tool-planning, cost-aware, tool-use, benchmark, 2411-16313

English Abstract

Utilizing large language models (LLMs) for tool planning has emerged as a promising avenue for developing general AI systems, where LLMs automatically schedule external tools (e.g., vision models) to tackle complex tasks based on task descriptions. To push this paradigm toward practical applications, it is crucial for LLMs to consider tool execution costs (e.g., execution time) for tool planning. Unfortunately, prior studies overlook the tool execution costs, leading to the generation of expensive plans whose costs outweigh their benefits in terms of task performance. To fill this gap, we propose the Cost-Aware Tool Planning with LLMs (CATP-LLM) framework, which for the first time provides a coherent design to empower LLMs for cost-aware tool planning. Specifically, To facilitate efficient concurrent tool execution and cost reduction, we design a tool planning language to enhance the LLM for creating multi-branch non-sequential plans. Moreover, we propose a cost-aware offline reinforcement learning algorithm to fine-tune the LLM to optimize the performance-cost trade-off in tool planning. In the lack of public cost-related datasets, we further present OpenCATP, the first dataset for cost-aware planning, which comprises 11,100 evaluation samples from diverse tasks. Extensive experiments show that CATP-LLM outperforms GPT-4 even when using Llama2-7B as its backbone, with the average improvement of 1.5%-93.9% in terms of plan quality. Codes and dataset are available at: this https URL.

English Summary

Obsidian Notes

内容获取路径：优先尝试 opencli arxiv paper 2411.16313 -f json，本轮遭遇 arXiv API HTTP 429 后，改由 arXiv 页面元数据与摘要回填。
中文导读与价值判断均锚定在条目已有摘要、论文摘要、作者、日期与分类信息上。
现代站点生成器按 content/{entry.id}.md 查找内容页；本文件写入 canonical content 目录，而不是 openclaw/content/。