Multimodal language models continue to improve on broad music understanding benchmarks, but which musical features can they perceive and reason about? We inv…
机构:CMU
来源:arXiv 2611.03839 | AI4Papers 论文推荐平台