Luca Ambrogioni (@LucaAmb) / X

Luca Ambrogioni

11K posts

Luca Ambrogioni

@LucaAmb

Ass. prof. of Machine Learning. PI of Generative Memory Lab (@DondersInst). Generative diffusion and statistical physics. AI realist.

Nijmegen, Nederland

Joined July 2011

Pinned
Luca Ambrogioni
@LucaAmb
May 11
1/?) As promised to Sander Dieleman (@sedielem), we’re finally excited to share: Towards Closing the Autoregressive Gap in Language Modeling via Entropy-Gated Continuous Bitstream Diffusion We show that continuous diffusion can achieve very strong language modeling performance
28K
Luca Ambrogioni
@LucaAmb
Apr 26, 2024
Diffusion models can be reformulated as equilibrium thermodynamical systems.
119K
Luca Ambrogioni
@LucaAmb
Nov 22, 2020
There are a handful of books that I re-read periodically every year and always learn something new. This is one of them. I highly recommend this to all ML/AI people who want to understand what statistics is.
Luca Ambrogioni
@LucaAmb
Mar 15, 2022
Deep learning works, symbolic models don't. It's that simple. If you want more symbolic models, then work hard and make them work. That's what NN people did, even when nobody believed in their research
Luca Ambrogioni
@LucaAmb
Sep 1, 2025
1/2) I am very happy to finally share something I have been working on and off for the past year: "The Information Dynamics of Generative Diffusion" This paper connects the entropy production, divergence of vector fields and spontaneous symmetry breaking in a unified framework
71K
Luca Ambrogioni
@LucaAmb
Oct 2, 2023
Modern Hopfield networks are related to transformers, but did you know that they are mathematically equivalent to generative diffusion models? Happy to share: "In search of dispersed memories: Generative diffusion models are associative memory networks" arxiv.org/abs/2309.17290
132K
Luca Ambrogioni
@LucaAmb
Oct 27, 2023
Happy to share: "The statistical thermodynamics of generative diffusion models" arxiv.org/abs/2310.17467 I describe diff. models in terms of Boltzmann distributions, order parameters and equations of state. hase transitions and critical scaling in the generative process!
88K
Luca Ambrogioni
@LucaAmb
Jun 24, 2024
Anthony Bonato
@Anthony_Bonato
Jun 23, 2024
What one thing in mathematics blew your mind when you first learned it?
53K
Luca Ambrogioni
@LucaAmb
Nov 27, 2024
FLow matching in a nutshell.
42K
Luca Ambrogioni
@LucaAmb
Sep 26, 2025
1/2) Have you have noticed that the forward process in a diffusion model looks a lot like the reparameterization trick in VAEs? It turns out that there is a deep connection! Curious? Watch our new vedio in the Generative Memory Lab channels (link below)
38K
Luca Ambrogioni
@LucaAmb
Apr 29, 2025
1/4) I am very happy to share our latest work on the information theory of generative diffusion: "Entropic Time Schedulers for Generative Diffusion Models" We find that the conditional entropy offers a natural data-dependent notion of time.
68K
Luca Ambrogioni
@LucaAmb
May 20, 2024
Did you know that the latent geometry of generative diffusion models can be extracted from the spectrum of eigenvalues of its Jacobi matrices? Paper: arxiv.org/pdf/2212.12611
49K
Luca Ambrogioni
@LucaAmb
Oct 9, 2024
The work of John Hopfield is 100% statistical physics. Calling him a 'computer scientist' is just ignorance.
Gabriel Peyré
@gabrielpeyre
Jul 11, 2023
Hopfield networks are recurrent networks minimizing an Ising-type energy parameterized by its weights. Learning weights means encoding patterns of +1/-1 as local minimizers. en.wikipedia.org/wiki/Hopfield_…
00:00
34K
Luca Ambrogioni
@LucaAmb
May 18, 2024
Memorization is generative diffusion models can be seen as a disordered phase transition!
57K