Picture for Luca Ambrogioni

Luca Ambrogioni

Dynamic Negative Guidance of Diffusion Models

Add code
Oct 18, 2024
Viaarxiv icon

Losing dimensions: Geometric memorization in generative diffusion

Add code
Oct 11, 2024
Figure 1 for Losing dimensions: Geometric memorization in generative diffusion
Figure 2 for Losing dimensions: Geometric memorization in generative diffusion
Figure 3 for Losing dimensions: Geometric memorization in generative diffusion
Figure 4 for Losing dimensions: Geometric memorization in generative diffusion
Viaarxiv icon

Manifolds, Random Matrices and Spectral Gaps: The geometric phases of generative diffusion

Add code
Oct 08, 2024
Viaarxiv icon

Robust and highly scalable estimation of directional couplings from time-shifted signals

Add code
Jun 04, 2024
Viaarxiv icon

The statistical thermodynamics of generative diffusion models

Add code
Oct 26, 2023
Viaarxiv icon

Stationarity without mean reversion: Improper Gaussian process regression and improper kernels

Add code
Oct 04, 2023
Viaarxiv icon

In search of dispersed memories: Generative diffusion models are associative memory networks

Add code
Sep 29, 2023
Viaarxiv icon

Spontaneous symmetry breaking in generative diffusion models

Add code
Jun 01, 2023
Viaarxiv icon

Closing the gap: Exact maximum likelihood training of generative autoencoders using invertible layers

Add code
May 19, 2022
Figure 1 for Closing the gap: Exact maximum likelihood training of generative autoencoders using invertible layers
Figure 2 for Closing the gap: Exact maximum likelihood training of generative autoencoders using invertible layers
Figure 3 for Closing the gap: Exact maximum likelihood training of generative autoencoders using invertible layers
Figure 4 for Closing the gap: Exact maximum likelihood training of generative autoencoders using invertible layers
Viaarxiv icon

Embedded-model flows: Combining the inductive biases of model-free deep learning and explicit probabilistic modeling

Add code
Oct 17, 2021
Figure 1 for Embedded-model flows: Combining the inductive biases of model-free deep learning and explicit probabilistic modeling
Figure 2 for Embedded-model flows: Combining the inductive biases of model-free deep learning and explicit probabilistic modeling
Figure 3 for Embedded-model flows: Combining the inductive biases of model-free deep learning and explicit probabilistic modeling
Figure 4 for Embedded-model flows: Combining the inductive biases of model-free deep learning and explicit probabilistic modeling
Viaarxiv icon