How do people go from interacting with a messy, nebulous world to honing in on relevant information and discovering underlying patterns to relate different experiences in some abstract world, often with just a few examples? Alessandro is interested in this computational problem, aspiring to develop algorithms with analogous capabilities that can generalize strongly to unknown problems with little experience (and priors). His previous research has explored alternative approaches to simplifying search in program induction by learning to continually restructure the program search space in a way that complements how it is guided.
Alessandro is interested in ideas from multiple domains, including probabilistic machine learning, theoretical computer science, and cognitive science
Alessandro recently received a BSc in Artificial Intelligence from The University of Edinburgh.