About

I am a postdoctoral researcher at the Institute for Logic, Language, and Computation at the University of Amsterdam. My interests lie at the intersection of natural language processing and probabilistic modelling. My research is part of the European UTTER project.

Current Interests

Selected Publications

An Approximate Sampler for Energy-based Models with Divergence Diagnostics
Bryan Eikema, Germán Kruszewski, Cristopher R Dance, Hady Elsahar, Marc Dymetman in Transactions on Machine Learning Research, 2022

Sampling-Based Approximations to Minimum Bayes Risk Decoding for Neural Machine Translation
Bryan Eikema and Wilker Aziz in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, 2022

Is MAP Decoding All You Need? The Inadequacy of the Mode in Neural Machine Translation
Bryan Eikema and Wilker Aziz in Proceedings of the 28th International Conference on Computational Linguistics (COLING), 2020 Best Paper Award

Talks

Projects

mbr-nmt: Sampling-based Minimum Bayes Risk decoding in Python.
AEVNMT.pt: A PyTorch-based framework for deep generative models of text.