A presentation by Lingfei Wu on graph-to-sequence learning models for NLP tasks, including neural translation and semantic parsing.