190 Thayer St.
Providence, RI 02912
I've now begun a postdoc with Thomas Serre at Brown University (DAAD fellowship). The information on this website still refers to my PhD studies in Edinburgh.
I am a PhD student at the Neuroinformatics DTC at the University of Edinburgh. Broadly speaking, I am doing research at the interface between computational neuroscience and machine learning: How does the brain learn and reason about the world? Can we translate insights about biological intelligence to the development of artificial intelligence? Can recent developments in machine learning inspire our approach to understanding the brain? My supervisors are Amos Storkey on the machine learning side, and Peggy Series on the computational neuroscience side.
In my PhD, I work on relating a model used in machine learning -- the Deep Boltzmann Machine -- to processing in the cerebral cortex. This type of model learns to represent sensory data without supervision. It provides a concrete neural network implementation of principles that are hypothesized to play an important role in the cortex, in particular, hierarchical Bayesian inference. To provide further evidence that these principles might apply to the brain, I aim to demonstrate how perceptual phenomena, from hallucinations to attentional processing, can be explained in this modelling framework.
Beyond my concrete topic of research, I am interested in a wide range of related issues, from the role of various brain structures in an evolutionary context, to philosophical and ethical implications of both neuroscience research and the development of AI.