Glow: Generative Flow with Invertible 1x1 Convolutions
Durk P. Kingma, Prafulla Dhariwal
–Neural Information Processing Systems
Flow-based generative models (Dinh et al., 2014) are conceptually attractive due to tractability of the exact log-likelihood, tractability of exact latent-variable inference,
Neural Information Processing Systems
Nov-20-2025, 20:13:55 GMT