用自然语言指导大模型路由，转化为约束优化问题。
LLM Routing as Reasoning: A MaxSAT View
- 将用户自然语言偏好转为硬约束和软约束
- 在25个模型基准上实现近可行推荐集
- 适合需要个性化模型选择的场景
将查询路由到合适的大型语言模型（LLM）极具挑战性，尤其当用户偏好以自然语言表达且模型属性仅部分可观测时。本文提出一种基于约束的语言条件化LLM路由解释，将其建模为加权最大满足性（MaxSAT/MaxSMT）问题，其中自然语言反馈在模型属性上施加硬约束与软约束。在此框架下，路由等价于选择能近似最大化反馈条件化子句满足度的模型组合。在包含25个模型的基准上进行的实证分析表明，基于语言反馈可生成近可行的推荐集合，而无反馈场景则暴露出系统性先验偏差。结果表明，LLM路由可在语言条件偏好下被理解为结构化的约束优化过程。
原文摘要 · Abstract (English)
Routing a query through an appropriate LLM is challenging, particularly when user preferences are expressed in natural language and model attributes are only partially observable. We propose a constraint-based interpretation of language-conditioned LLM routing, formulating it as a weighted MaxSAT/MaxSMT problem in which natural language feedback induces hard and soft constraints over model attributes. Under this view, routing corresponds to selecting models that approximately maximize satisfaction of feedback-conditioned clauses. Empirical analysis on a 25-model benchmark shows that language feedback produces near-feasible recommendation sets, while no-feedback scenarios reveal systematic priors. Our results suggest that LLM routing can be understood as structured constraint optimization under language-conditioned preferences.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。