Audio-Aware Large Language Models as Judges for Speaking Styles Paper β’ 2506.05984 β’ Published Jun 6, 2025 β’ 15