Preprint
Sep 2026
Do Audio Representations Compose Additively?
Compositionality, the ability to represent complex acoustic scenes as combinations of simpler sound sources, is central to auditory perception and classical additive signal models. Still, it remains unclear whether modern pre-trained audio representations internalize additive structure without compositional supervision...
Chen-Hao Xue, Zhi-Jin Guo, Joyraj Chakraborty et al.
· 0 citations