Large language model (LLM) routing distributes inference work across different models, advancing the cost-quality Pareto frontier of LLM serving. While coars…
机构:清华
来源:arXiv 2610.12242 | AI4Papers 论文推荐平台