1 paper
Yifei Li, Zihui Gao, Laks V. S. Lakshmanan
Large language models (LLMs) achieve impressive performance across multiple domains, but using the most capable model for every query is prohibitive at scale. LLM routing exploits…