My current research involves building deep generative models for natural language, particularly in the context of dialogue systems. I’m generally interested in deep learning, reinforcement learning, causal models, and biologically plausible AI.

Recently, I’ve been working on building dialogue models and datasets. I’m also currently working on building a neural network-based model for evaluating dialogue responses in a way that actually correlates with human judgement, which we recently found was not the case for existing dialogue metrics.