#disentanglement

A β-VAE maximises $\mathbb{E}_{q(z \mid x)}[\log p(x \mid z)] - \beta\, D_{\mathrm{KL}}(q(z \mid x) \,\|\, p(z))$ with an isotropic unit Gaussian prior $p(z)$. True or false: raising $\beta$ from 1 to, say, 4 typically gives both more disentangled latent factors and more accurate reconstructions.