Picture for Pascal Vincent

Pascal Vincent

Universite de Montreal

Compositional Risk Minimization

Add code
Oct 08, 2024
Viaarxiv icon

WorldSense: A Synthetic Benchmark for Grounded Reasoning in Large Language Models

Add code
Nov 27, 2023
Viaarxiv icon

Self-Supervised Disentanglement by Leveraging Structure in Data Augmentations

Add code
Nov 15, 2023
Figure 1 for Self-Supervised Disentanglement by Leveraging Structure in Data Augmentations
Figure 2 for Self-Supervised Disentanglement by Leveraging Structure in Data Augmentations
Figure 3 for Self-Supervised Disentanglement by Leveraging Structure in Data Augmentations
Figure 4 for Self-Supervised Disentanglement by Leveraging Structure in Data Augmentations
Viaarxiv icon

Motif: Intrinsic Motivation from Artificial Intelligence Feedback

Add code
Sep 29, 2023
Viaarxiv icon

Discovering environments with XRM

Add code
Sep 28, 2023
Figure 1 for Discovering environments with XRM
Figure 2 for Discovering environments with XRM
Figure 3 for Discovering environments with XRM
Figure 4 for Discovering environments with XRM
Viaarxiv icon

PUG: Photorealistic and Semantically Controllable Synthetic Data for Representation Learning

Add code
Aug 08, 2023
Figure 1 for PUG: Photorealistic and Semantically Controllable Synthetic Data for Representation Learning
Figure 2 for PUG: Photorealistic and Semantically Controllable Synthetic Data for Representation Learning
Figure 3 for PUG: Photorealistic and Semantically Controllable Synthetic Data for Representation Learning
Figure 4 for PUG: Photorealistic and Semantically Controllable Synthetic Data for Representation Learning
Viaarxiv icon

Predicting masked tokens in stochastic locations improves masked image modeling

Add code
Jul 31, 2023
Figure 1 for Predicting masked tokens in stochastic locations improves masked image modeling
Figure 2 for Predicting masked tokens in stochastic locations improves masked image modeling
Figure 3 for Predicting masked tokens in stochastic locations improves masked image modeling
Figure 4 for Predicting masked tokens in stochastic locations improves masked image modeling
Viaarxiv icon

Identifiability of Discretized Latent Coordinate Systems via Density Landmarks Detection

Add code
Jun 28, 2023
Viaarxiv icon

Do SSL Models Have Déjà Vu? A Case of Unintended Memorization in Self-supervised Learning

Add code
Apr 28, 2023
Figure 1 for Do SSL Models Have Déjà Vu? A Case of Unintended Memorization in Self-supervised Learning
Figure 2 for Do SSL Models Have Déjà Vu? A Case of Unintended Memorization in Self-supervised Learning
Figure 3 for Do SSL Models Have Déjà Vu? A Case of Unintended Memorization in Self-supervised Learning
Figure 4 for Do SSL Models Have Déjà Vu? A Case of Unintended Memorization in Self-supervised Learning
Viaarxiv icon

Objectives Matter: Understanding the Impact of Self-Supervised Objectives on Vision Transformer Representations

Add code
Apr 25, 2023
Viaarxiv icon