Friday 31 January 2025, 11am - Martha Lewis

Abstract: Recent neural approaches to modelling language and concepts have proven quite effective, with a proliferation of large models trained on correspondingly massive datasets. However, these models still fail on some tasks that humans, and symbolic approaches, can easily solve. Large neural models are also, to a certain extent, black boxes - particularly those that are proprietary. There is therefore a need to integrate compositional and neural approaches, firstly to potentially improve the performance of large neural models, and secondly to analyze and explain the representations that these systems are using. In this talk I will present results showing that large neural models can fail at tasks that humans are able to do, and discuss alternative, theory-based approaches that have the potential to perform more strongly. I will give applications in language, reasoning, and vision. Finally, I will present some future directions in understanding the types of reasoning or symbol manipulation that large neural models may be performing

 

Bio: Martha Lewis is Assistant Professor of Neurosymbolic AI in the Institute of Logic, Language, and Computation, University of Amsterdam. Before ILLC, she was most recently a Lecturer in the School of Engineering Mathematics and Technology at Bristol University. She held a Veni fellowship at the ILLC at the University of Amsterdam, was a postdoc in the Quantum Group in the Department of Computer Science, University of Oxford, and did her PhD at the University of Bristol, in the Bristol Centre for Complexity Sciences.