1
MusTBENCH: Benchmarking and Advancing Temporal Grounding in Music LLMs
音乐大模型时序理解新基准MusTBENCH，填补音乐AI时序评估空白，推动标准化研究。
arXiv:2605.29300v1 Announce Type: cross Abstract: Recent Large Audio-Language Models (LALMs) have demonstrated promising abilities in understanding mu…