LLM委托合约：技术与努力选择中的道德风险

AI 摘要

该论文将经典委托-代理框架扩展至LLM代理选择技术与努力水平（如token预算）的场景，建模了凹性饱和生产函数，推导了最优线性合约，并基于MATH和MMLUPro基准校准模型，证明简单线性合约能有效激励技术感知的委托。

正文节选

Contracting for LLM Delegation: Moral Hazard in Technology and Effort Choice Abstract We extend the standard Principal-Agent framework to scenarios where the Agent selects from a suite of technologies, each characterized by a distinct cost-capability profile. This framework is increasingly critical in the era of Large Language Models (LLMs), where Agents choose both a model and an associated effort level (e.g., token budget). We model the relationship between output quality and effort as a conca

发布时间：2026-08-20 12:00

抓取时间：2026-08-20 12:22

来源机构：arXiv