- Machine learning, in particular deep and reinforcement/bandit learning for NLP
- Machine translation and quality estimation.
- Attending the Deep Learning and Reinforcement Learning Summer School 2018 in Toronto, presenting our ACL18 paper (slides).
- The NLP Highlights podcast is discussing our EMNLP17 paper about Easy-First Decoding with André Martins. Listen here.
- Paper accepted at ACL: "Reliability and Learnability of Human Bandit Feedback for Sequence-to-Sequence Reinforcement Learning" J. Kreutzer, J. Uyheng, S. Riezler. We released the data here.