Director of the Computational Linguistics Postgraduate Programme · UCL PaLS
Member of the European Laboratory for Learning and Intelligent Systems · UCL ELLIS Unit
Co-founder and Scientific Director · Parallax
My research explores the information processing principles that underly the ability to use and learn language — both in human and artificial language processing systems. I'm also increasingly interested in biological and artificial cognitive systems more broadly, and by extension to extra-linguistic aspects of perception, action, and interaction. I am committed to advancing the science of AI evaluation, with a focus on language as well as broader aspects of agency and safety. In 2018, I co-authored a paper that, according to the very kind Aaron Mueller, introduced the first causal (mechanistic?) interpretability method for language models.
- 2025Senior Research Scientist, UK AI Security Institute, Science of Evaluation & Testing Teams
- 2023–25Postdoctoral Fellow, ETH Zurich, Department of Computer Science, working with Ryan Cotterell
- 2019–23PhD, University of Amsterdam, Institute for Logic, Language and Computation, advised by Raquel Fernández
UCL is recruiting a Lecturer in Computational Linguistics for a fixed-term two-year position with substantial time for research.
I'm looking for prospective PhD students interested in AI evaluation, auditing and interpretability, or computational models of human language use. I have funding for one position at UCL, recruited through the ELLIS PhD Programme. Applications close on 31 October 2026. The strongest candidates may also be put forward for other funding programmes.
Three papers at ICML 2026: one combining behavioural and representational analyses to evaluate goal-directedness in LLM agents; one introducing a statistical framework for measuring bias and uncertainty in LLM-as-judge evaluations; and one showing that reasoning aligns the cognitive profiles of LLMs with those of humans (presented at the Mechanistic Interpretability and Combining Theory and Benchmarks workshops).
Two papers presented at ACL 2026 in San Diego. One paper, Probing for Reading Times, studies whether language model representations contain signatures of human reading times. The other, Surprisal Minimisation over Goal-directed Alternatives Predicts Production Choice in Dialogue, uses a new approach to generating open-ended goal-agnostic and goal-oriented alternatives to study speaker choice sensitivity to various information-theoretic notions of comprehension and production cost.
We have announced the call for papers for the LM Playschool Workshop and Challenge. We invite submissions exploring the frontier of language agents that learn, adapt, and improve through situated interaction, with a focus on conversational, collaborative, goal-oriented, and multi-turn environments.