1 paper
Yuanjian Chen, Yang Xiao, Han Yin +3
Large Audio Language Models (LALMs) are increasingly capable of reasoning over audio, yet existing benchmarks offer limited coverage of reasoning in polyphonic audio, where multipl…