Abstract

In this three part talk, I will present some of our recent efforts that aim to control and adapt neural models to work more effectively in end applications. The first part will focus on how to repurpose pre-trained neural entailment models for multi-hop QA, some ongoing work on creating hoppy datasets that ensure multi-hop reasoning, and decomposing large QA models to run on mobile devices. In the second part, I will present methods for learning structured latent spaces for better control in modeling and generating event sequences. In the third part, I will talk briefly about modeling target side syntax for machine translation.

Biography

Niranjan is an Assistant Professor in the Computer Science department at Stony Brook University, where he heads the LUNR lab, short for Language Understanding and Reasoning lab. LUNR lab focuses on many problems in NLP including QA, event knowledge, language generation, and efficient NLP for mobile devices. Prior to joining Stony Brook, he was a post-doctoral researcher in the University of Washington, and was one of the early members of the Allen Institute for Artificial Intelligence. Niranjan completed his PhD in Computer Science from the University of Massachusetts Amherst. He likes food and wine just enough that he sometimes mentions it as part of his bio.