I do research in the intersection of formal language theory and natural language processing (NLP). In particular, I'm investigating the properties and expressive power of neural network architectures by relating them to discrete computing devices, such as finite automata. A question of interest to me is how these properties can be studied through the modular structure of the network, i.e., how they are affected by the interplay among individual components.
In general, any application of logic or formal methods to NLP will grab my attention.
