1 paper · 1 filter
Zhanliang Wang, Jiancong Xiao, Ruochen Jin +3
Calibration measures whether a model's predicted confidence aligns with its empirical accuracy, and is central to the reliable deployment of large language models (LLMs) in high-st…