1 paper across 1 session
Latent recalibration learns a radial transform that calibrates normalizing flows, preserving an explicit PDF and improving NLL.
