Music Hallucination in Audio-Language Models: A Hierarchical Formulation and Empirical Study
Audio-language models increasingly generate confident music descriptions that are unsupported by the input audio. We present, to our knowledge, the first music-specific, layer-wise, multi-paradigm empirical study of hallucination in audio-language models and formulate it as a hierarchical perceptual grounding failure a...
Yu Liu, Jia-Hui Liu, Zhi-Lin Liu et al.
· 0 citations