My interests lie in the intersection of machine learning, natural language processing and vision to understand and develop human-like artificial intelligence. I’m especially interested in multimodal and multilingual grounding and generation, and in probing the abilities and properties of such models.