Human Evaluation of Conversations is an Open Problem: Comparing the Sensitivity of Various Methods for Evaluating Dialogue Agents
Eric Smith, Orion Hsu, Rebecca Qian, Stephen Roller, Y-Lan Boureau, Jason Weston
People
Stephen Roller
Research Engineer
I am a Research Engineer at Meta AI Research. Previously, I completed my PhD at the University of Texas at Austin, where I studied applications of word embeddings in areas including textual entailment, lexical entailment, lexical relations, paraphrasing, and non-textual modalities. I was also briefly a visiting student at The University of Stuttgart, and completed my undergraduate studies at North Carolina State University.
Interests
Natural language processing, semantics, word embeddings, deep learning, and lexical entailment.
Latest Publications
NLP for Conversational AI Workshop at ACL - 2022
Wizard of Wikipedia: Knowledge-Powered Conversational Agents
Emily Dinan, Stephen Roller, Kurt Shuster, Angela Fan, Michael Auli, Jason Weston
Not All Memories are Created Equal: Learning to Forget by Expiring
Sainbayar Sukhbaatar, Spencer Poff, Stephen Roller, Arthur Szlam, Jason Weston, Angela Fan
The Dialogue Dodecathlon: Open-Domain Knowledge and Image Grounded Conversational Agents
Kurt Shuster, Da Ju, Stephen Roller, Emily Dinan, Y-Lan Boureau, Jason Weston
Don’t Say That! Making Inconsistent Dialogue Unlikely with Unlikelihood Training
Margaret Li, Stephen Roller, Ilia Kulikov, Sean Welleck, Y-Lan Boureau, Kyunghyun Cho, Jason Weston
Inferring Concept Hierarchies from Text Corpora via Hyperbolic Embeddings
Matt Le, Stephen Roller, Laetitia Papaxanthos, Douwe Kiela, Maximilian Nickel