1 paper
Zhaoyang Zhang, Run Shao, Dongyue Wu +4
Understanding why independently trained neural networks from different modalities converge toward shared representations, and where this convergence leads, remains an open question…