arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.11574cs.AI
收藏

学习配置代理AI系统

Learning to Configure Agentic AI Systems

School of Computing and Augmented Intelligence(计算与增强智能学院);Arizona State University(亚利桑那州立大学)

Aditya Taparia, Som Sagar, Ransalu Senanayake

AI总结:

本文提出了一种基于半马尔可夫决策过程(SMDP)的代理配置方法,通过ARC模型动态选择查询特定的代理配置,从而在多个基准测试中提升了推理准确性、工具使用准确性和τ-Bench(Airline)Pass的成功率。

摘要:

配置基于LLM的代理系统涉及从庞大的组合设计空间中选择工作流、工具、令牌预算和提示,而目前通常通过固定的模板或手工调整的启发式方法处理,这些方法无论查询难度如何都应用相同的配置,导致行为脆弱和计算浪费。为了解决这个问题,我们将代理配置建模为半马尔可夫决策过程(SMDP),其中每个配置都是一种时间扩展的选项,决定了代理系统如何处理查询,并引入了ARC(Agentic Resource & Configuration learner),一种轻量级的分层策略,能够动态选择查询特定的代理配置。在推理、工具使用和代理基准测试中,ARC在与预算匹配的工具增强LLM相比,平均推理准确性提高了31.3%,工具使用准确性提高了13.95%,并将τ-Bench(Airline)Pass的成功率从9.0%提升到18.0%。这些结果表明,学习查询特定的代理配置是“一刀切”设计的一种强大替代方案。

英文摘要:

Configuring LLM-based agent systems involves choosing workflows, tools, token budgets, and prompts from a large combinatorial design space, and is typically handled today by fixed templates or hand-tuned heuristics that apply the same configuration regardless of query difficulty, leading to brittle behavior and wasted compute. To address this, we formulate agent configuration as a semi-Markov decision process (SMDP) where each configuration acts as a temporally extended option that determines how an agent system processes a query, and introduce introduce ARC (Agentic Resource & Configuration learner), a lightweight hierarchical policy that dynamically selects query-specific agent configurations. Across reasoning, tool-use, and agentic benchmarks, ARC consistently improves over budget-matched tool-augmented LLMs, increasing average reasoning accuracy by 31.3%, tool-use accuracy by 13.95%, and doubling τ-Bench (Airline) Pass^1 success from 9.0% to 18.0%. These results demonstrate that learning per-query agent configurations is a powerful alternative to "one size fits all" designs.

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。

↑