arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.16653cs.AI
收藏

Agent技能框架:中小型语言模型在工业环境中的潜力视角

Agent Skill Framework: Perspectives on the Potential of Small to Medium Language Models in Industrial Environments

University of Luxembourg(卢森堡大学);Foyer S.A.(Foyer公司);Princeton University(普林斯顿大学);Université Paris-Saclay(巴黎萨克雷大学)

Yangjie Xu, Lujun Li, Lama Sleem, Niccolo Gentile, Yewei Song, Yiqun Wang, Siming Ji, Wenbo Wu, Hui Huang, Radu State

AI总结:

研究在资源受限的工业场景中,中小型开源语言模型(270M-80B)使用Agent技能的效果,发现30B-80B模型受益显著,而小型模型技能选择困难,思考变体提升有限且增加GPU开销。

摘要:

Agent技能被主要的Agent框架广泛支持,并在专有模型上表现良好,但其对中小型开源语言模型(270M-80B)的有效性尚未充分探索。我们系统研究了资源受限工业环境中的技能范式,在这些环境中,由于数据安全和预算限制,依赖专有API不切实际。在两个开源任务和一个真实保险索赔分类任务中,我们发现非常小的模型在可靠技能选择上存在困难,而30B-80B左右的模型则显著受益。思考变体未显示出技能带来的显著改进,同时由于过度思考导致GPU使用增加。这些发现揭示了GPU成本与Agent性能之间的权衡,并为现实环境中有效的技能配置和中小型语言模型部署提供了可操作的见解。

英文摘要:

Skills are getting increasingly used in recent agentic frameworks, performing robustly with proprietary LLMs. Yet, their effectiveness for small and medium-sized open-source models (270M--80B) remains underexplored, especially under industrial, resource-constrained conditions. As such, we present a large-scale empirical study evaluating the skill paradigm's feasibility across two open-source classification tasks: one insurance-claim task based on proprietary data, and a set of Python programming problems. We find that very small models frequently fail to retrieve or select correct skills, substantially limiting their usefulness in the evaluated skill-augmented settings, while larger 30B--80B models gain substantial benefits from skill augmentation. Furthermore, we provide a deployment-oriented GPU profile, reporting VRAM time occupancy (GB$\cdot$min) across models under our evaluation settings---a deployment-oriented metric that may help characterize GPU-memory costs, where small and medium models are typically preferred for VRAM control. These results characterize how the benefits of skill augmentation vary with model scale, task complexity, and GPU budget, and provide deployment-oriented observations.

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。

↑