How Can NLP Help Reinforcement Learning?

How Can NLP Help Reinforcement Learning? (via Zoom)

Abstract: In recent years, reinforcement learning (RL) has seen considerable success in games and robotics as well as NLP applications like dialog systems or text generation. However, the question of what language can provide for RL remains relatively under-explored. In this talk, I make the case for leveraging language understanding in developing interactive agents that can scale to multiple tasks and operate in a wider range of scenarios beyond the ones they are trained on. Natural language allows us to incorporate semantic structure into the RL framework while also making it easier to obtain guidance from humans. Specifically, I will show how several parts of the RL setup (e.g. transitions, rewards, actions, constraints) can be expressed in language to build agents that can intelligently explore combinatorially large spaces, generalize to unseen environments and learn in a safe manner.

Bio: Karthik Narasimhan is an assistant professor in the Computer Science department at Princeton University. His research spans the areas of natural language processing and reinforcement learning, with the goal of building intelligent agents that learn to operate in the world through both their own experience and leveraging existing human knowledge. Karthik received his PhD from MIT in 2017, and spent a year as a research scientist at OpenAI prior to joining Princeton in 2018. His work has received a best paper award at EMNLP 2016 and an honorable mention for best paper at EMNLP 2015.