-
) information-theoretic active learning, and c) capturing uncertainty in deep learning models (including large language models). The successful postholder will hold or be close to the completion of a PhD/DPhil in
-
for solving problems related to safe robotic exploration and planning under epistemic uncertainty. We typically work with Markov decision processes (MDPs), and are looking for someone with experience in both
-
• Uncertainty quantification around LLMs • Constrained optimal experimental design (active learning) • Combining models and combining data / Realistic simulation of clinical trials • Developing
-
, the centre will initially focus on some of the following thematic areas: • Decision analysis under model misspecification • Uncertainty quantification around LLMs • Constrained optimal