We use computational models and psychological experiments to understand human cognition, and build systems that take inspiration from human cognition, with a particular emphasis on the ability to learn and act in new and changing situations.
We are accepting new Ph.D. students – see how to contact us.
|Nov 2, 2023||Edinburgh has two new cogsci-related Centres for Doctoral Training (CDTs): “Responsible and Trustworthy in-the-world NLP”, and “Dependable and Deployable AI for Robotics”. These are successors to CDTs that have funded PhD students in the group. Here’s the anncouncement from Informatics.|
|Nov 1, 2023||Verna Dankers’s “Non-Compositionality in Sentiment: New Data and Analyses” has been accepted at EMNLP Findings and will be presented at the BlackBoxNLP workshop.|
|Oct 27, 2023||Jan-Philipp Fränken’s “Naïve information aggregation in human social learning” is in press in Cognition.|
|Sep 28, 2023||Chentian will be presenting “Learning How to Infer Partial MDPs for In-Context Adaptation and Exploration” – based on work from her recent internship at DeepMind – at the NeurIPS WiML workshop.|
|Aug 7, 2023||Bonan Zhao’s “How cognition bootstraps its way to complex concepts” has been accepted for publication at Nature Human Behaviour.|