Lucas Souza discusses few-shot learning and relevant benchmarks used to measure performance in these settings. This is part of a series of research meetings aimed at reviewing common training paradigms and benchmarks in machine learning, and their relation to the Thousand Brains Theory of Intelligence.