AI RESEARCH
Training-Free Vector Quantization via Gaussian VAEs
arXiv CS.LG
•
ArXi:2512.06609v3 Announce Type: replace Vector-quantized variational autoencoders (VQ-VAEs) are discrete autoencoders that compress images into discrete tokens. However, they are difficult to train due to discretization. In this paper, we propose a simple yet effective technique dubbed Gaussian Quant (GQ), which first trains a Gaussian VAE under certain constraints and then converts it into a VQ-VAE without additional