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 2602.11931cs.CLcs.AI
收藏

AdaptEvolve: 通过自适应模型选择提高进化AI代理的效率

AdaptEvolve: Improving Efficiency of Evolutionary AI Agents through Adaptive Model Selection

IIT Kharagpur(印度理工学院Kharagpur分校);Advanced Micro Devices, Inc. (AMD)(先进微器件公司(AMD))

Pretam Ray, Pratik Prabhanjan Brahma, Zicheng Liu, Emad Barsoum

AI总结:

本文提出AdaptEvolve方法,通过自适应选择大语言模型提升进化代理效率,实验证明在保持准确性的同时降低推理成本37.9%。

摘要:

进化代理系统通过反复调用大型语言模型(LLMs)进行推理,加剧了计算效率与推理能力之间的权衡。本文引入AdaptEvolve:一种基于内在生成信心的自适应LLM选择方法,用于多LLM进化细化框架中。实验证明,信心驱动的选择在保持静态大模型基线的上界准确性的同时,减少了总推理成本37.9%。代码可在https://github.com/raypretam/adaptive_llm_selection获取。

英文摘要:

Evolutionary agentic systems intensify the trade-off between computational efficiency and reasoning capability by repeatedly invoking large language models (LLMs) during inference. This setting raises a central question: how can an agent dynamically select an LLM that is sufficiently capable for the current generation step while remaining computationally efficient? While model cascades offer a practical mechanism for balancing this trade-off, existing routing strategies typically rely on static heuristics or external controllers and do not explicitly account for model uncertainty. We introduce AdaptEvolve: Adaptive LLM Selection for Multi-LLM Evolutionary Refinement within an evolutionary sequential refinement framework that leverages intrinsic generation confidence to estimate real-time solvability. Empirical results show that confidence-driven selection yields a favourable Pareto frontier, reducing total inference cost by an average of 37.9% across benchmarks while retaining 97.5% of the upper-bound accuracy of static large-model baselines. Our code is available at https://github.com/raypretam/adaptive_llm_selection.

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。

↑