Matthijs T. J. Spaan

Contextual Similarity Distillation: Ensemble Uncertainties with a Single Model

Universal Value-Function Uncertainties

COP-Q: Safety-First Reinforcement Learning with Cholesky Ordered Projection

Exploration Implies Data Augmentation: Generalisation in Contextual MDPs

Parallelizing Tree Search With Twice Sequential Monte Carlo

Off-Policy Safe Reinforcement Learning with Cost-Constrained Optimistic Exploration

How Ensembles of Distilled Policies Improve Generalisation in Reinforcement Learning

Value Improved Actor Critic Algorithms

Epistemic Monte Carlo Tree Search

Trust-Region Twisted Policy Improvement

Training on more Reachable Tasks for Generalisation in Reinforcement Learning