What is a language agent? An AI system that takes advantage of information latent to textual observations (and encoded in an LLM) to take actions in an environment? Or an agent whose action space necessarily includes the use of language with other language-using agents? In this talk, I will discuss some recent works that take different perspectives on what it means to be a "language agent" sharing a world with the original (and perhaps only?) natural language agents -- humans.
Alane Suhr recently joined EECS and BAIR at UC Berkeley as an Assistant Professor. Alane's work focuses on building language-using systems that communicate with and learn from human users in collaborative, situated interactions. Prior to joining Berkeley, Alane completed a PhD in Computer Science at Cornell University / Cornell Tech and spent a year afterwards as a Young Investigator at the Allen Institute for AI.