D
Dzmitry Bahdanau
Researcher at McGill University
Publications - 61
Citations - 72393
Dzmitry Bahdanau is an academic researcher from McGill University. The author has contributed to research in topics: Computer science & Artificial neural network. The author has an hindex of 26, co-authored 53 publications receiving 58851 citations. Previous affiliations of Dzmitry Bahdanau include Université de Montréal & Jacobs University Bremen.
Papers
More filters
Posted Content
Automated curriculum generation for Policy Gradients from Demonstrations
TL;DR: A training curriculum is developed that uses a nominal number of expert demonstrations and trains the agent in a manner that draws parallels from one of the ways in which humans learn to perform complex tasks, i.e by starting from the goal and working backwards.
Dissertation
On sample efficiency and systematic generalization of grounded language understanding with deep learning
Posted Content
Compositional Generalization in Dependency Parsing
TL;DR: This article introduced a set of dependency parses for CFQ, and used this to analyze the behavior of a state-of-the-art dependency parser (Qi et al., 2020) on the CFQ dataset.
Posted Content
Understanding by Understanding Not: Modeling Negation in Language Models
TL;DR: This paper proposed to augment the language modeling objective with an unlikelihood objective that is based on negated generic sentences from a raw text corpus, which reduced the mean top-1 error rate to 4% on the negated LAMA dataset.
Journal ArticleDOI
Combating False Negatives in Adversarial Imitation Learning (Student Abstract)
Konrad Żołna,Chitwan Saharia,Leonard Boussioux,David Yu-Tung Hui,Maxime Chevalier-Boisvert,Dzmitry Bahdanau,Yoshua Bengio +6 more
TL;DR: The False Negatives problem is defined and a method that solves the problem by leveraging the nature of goal-conditioned tasks is proposed, dubbed Fake Conditioning, which improves sample efficiency over the baselines by at least an order of magnitude.