@article{
wu2026from,
title={From Deferral to Learning: Online In-Context Knowledge Distillation for {LLM} Cascades},
author={Yu Wu and Shuo Wu and Ye Tao and Yansong Li and Anand D. Sarwate},
journal={Transactions on Machine Learning Research},
issn={2835-8856},
year={2026},
url={https://openreview.net/forum?id=KUcCF5qqFL},
note={}
}