Picture for Fredrik Lindsten

Fredrik Lindsten

LIENS, INRIA Paris - Rocquencourt, MSR - INRIA

On Partial Prototype Collapse in the DINO Family of Self-Supervised Methods

Add code
Oct 17, 2024
Viaarxiv icon

Continuous Ensemble Weather Forecasting with Diffusion models

Add code
Oct 07, 2024
Viaarxiv icon

Prior Learning in Introspective VAEs

Add code
Aug 25, 2024
Viaarxiv icon

Towards understanding epoch-wise double descent in two-layer linear neural networks

Add code
Jul 13, 2024
Viaarxiv icon

Probabilistic Weather Forecasting with Hierarchical Graph Neural Networks

Add code
Jun 07, 2024
Viaarxiv icon

DINO as a von Mises-Fisher mixture model

Add code
May 17, 2024
Figure 1 for DINO as a von Mises-Fisher mixture model
Figure 2 for DINO as a von Mises-Fisher mixture model
Figure 3 for DINO as a von Mises-Fisher mixture model
Figure 4 for DINO as a von Mises-Fisher mixture model
Viaarxiv icon

On the connection between Noise-Contrastive Estimation and Contrastive Divergence

Add code
Feb 26, 2024
Figure 1 for On the connection between Noise-Contrastive Estimation and Contrastive Divergence
Figure 2 for On the connection between Noise-Contrastive Estimation and Contrastive Divergence
Figure 3 for On the connection between Noise-Contrastive Estimation and Contrastive Divergence
Figure 4 for On the connection between Noise-Contrastive Estimation and Contrastive Divergence
Viaarxiv icon

Discriminator Guidance for Autoregressive Diffusion Models

Add code
Oct 24, 2023
Viaarxiv icon

Graph-based Neural Weather Prediction for Limited Area Modeling

Add code
Sep 29, 2023
Viaarxiv icon

Temporal Graph Neural Networks for Irregular Data

Add code
Feb 16, 2023
Viaarxiv icon