Tinne Tuytelaars

Co-LoRA: Collaborative Model Personalization on Heterogeneous Multi-Modal Clients

Distributional value gradients for stochastic environments

Sliced Distributional Reinforcement Learning

Forgetting Order of Continual Learning: What is Learned First is Forgotten Last

A Simple Framework for Open-Vocabulary Zero-Shot Segmentation

Object-Centric Pretraining via Target Encoder Bootstrapping

Failure-Proof Non-Contrastive Self-Supervised Learning

Collapse-Proof Non-Contrastive Self-Supervised Learning

Predicting the Susceptibility of Examples to Catastrophic Forgetting

Minimizing Dependence between Embedding Dimensions with Adversarial Networks

Same Accuracy, Twice As Fast: Continual Learning Surpasses Retraining From Scratch

Distributional Sobolev reinforcement learning