| Location: | London, Hybrid |
|---|---|
| Salary: | £45,103 to £51,246 |
| Hours: | Full Time |
| Contract Type: | Fixed-Term/Contract |
| Placed On: | 3rd July 2026 |
|---|---|
| Closes: | 26th July 2026 |
| Job Ref: | B02-10829 |
An opportunity has arisen for a Research Fellow in Machine Cognition in UCL Linguistics, within the Division of Psychology and Language Sciences and Faculty of Brain Sciences. UCL is ranked eighth in the QS World University Rankings; the Division is ranked fifth globally for psychology in QS 2026, and UCL is first in the UK for research power in psychology, psychiatry and neuroscience in REF 2021. The postholder will join Mario Giulianelli’s research group, which studies information processing in human and artificial systems. A core part of the group’s work also concerns the evaluation, interpretability and cognitive modelling of language models. The group is part of the UCL ELLIS Unit and has links to the UCL AI Centre and UCL Computer Science.
About the role
The postholder will carry out original research on machine cognition, focusing on how language model agents and related artificial neural network systems represent beliefs and goals. The project will investigate whether such systems represent uncertainty over environment states and preferences over possible states, and whether these representations can be reliably extracted, evaluated and manipulated. The role combines interpretability methods, behavioural evaluation, probabilistic modelling and ideas from cognitive science, neuroscience and theories of learning. The broader aim is to develop methods that support high-confidence claims about the goals pursued and beliefs held by artificial agents. Dedicated resources are available for travel, equipment, compute and API access.
About you
We seek a candidate with a PhD, or equivalent experience, in computational linguistics, artificial intelligence, cognitive science, neuroscience, computer science, psychology or a related field. They should have strong Python skills, experience with transformer language models, knowledge of interpretability methods and experience with probabilistic or computational modelling. Experience with language model evaluation, cognitive modelling, reinforcement learning, goal-directed behaviour, learning theory or large-scale GPU/cluster workflows would be advantageous.
Further Details
The advert will close on 26 July 2026 at 23:59 GMT. In the event we get a high number of applications, we may close the advert early before the published closing date. Early application submission is recommended.
Customer advert reference: B02-10829
Type / Role:
Subject Area(s):
Location(s):