Search Results for author:
Found 8 papers, 4 papers with code
Diverse capability and scaling of diffusion and auto-regressive models when learning abstract rules
We evaluated their ability to generate structurally consistent samples and perform panel completion via unconditional and conditional sampling.
The Hidden Linear Structure in Score-Based Models and its Application
We claimed that for well-trained diffusion models, the learned score at a high noise scale is well approximated by the linear score of Gaussian.
Diffusion Models Generate Images Like Painters: an Analytical Theory of Outline First, Details Later
How do diffusion generative models convert pure noise into meaningful images?
On the Level Sets and Invariance of Neural Tuning Landscapes
This shows the power of level sets as a conceptual tool to understand neuronal activations over image space.
High-performance Evolutionary Algorithms for Online Neuron Control
To find these patterns, we have used black-box optimizers to search a 4096d image space, leading to the evolution of images that maximize neuronal responses.
On the use of Cortical Magnification and Saccades as Biological Proxies for Data Augmentation
Critically, we find that random cropping can be substituted by cortical magnification, and saccade-like sampling of the image could also assist the representation learning.
The Geometry of Deep Generative Image Models and its Applications
Our results illustrate that defining the geometry of the GAN image manifold can serve as a general framework for understanding GANs.
A Geometric Analysis of Deep Generative Image Models and Its Applications
We show that the use of this metric allows for more efficient optimization in the latent space (e. g. GAN inversion) and facilitates unsupervised discovery of interpretable axes.
