Neural encoder-decoder models for language generation can be trained to predict words
directly from linguistic or non-linguistic inputs. When generating with these so-called end-to …

Quite surprisingly, exact maximum a posteriori (MAP) decoding of neural language
generators frequently leads to low-quality results. Rather, most state-of-the-art results on …

Solving math word problems requires deductive reasoning over the quantities in the text.
Various recent research efforts mostly relied on sequence-to-sequence or sequence-to-tree …

Drug discovery for a protein target is a very laborious, long and costly process. Machine
learning approaches and, in particular, deep generative networks can substantially reduce …

When generating text from probabilistic models, the chosen decoding strategy has a
profound effect on the resulting text. Yet the properties elicited by various decoding …

Recent studies have revealed a number of pathologies of neural machine translation (NMT)
systems. Hypotheses explaining these mostly suggest there is something fundamentally …

In this paper, we present a substantial step in better understanding the SOTA sequence-to-
sequence (Seq2Seq) pretraining for neural machine translation~(NMT). We focus on …

Beam search is the go-to method for decoding auto-regressive machine translation models.
While it yields consistent improvements in terms of BLEU, it is only concerned with finding …

We propose an alternate approach to quantifying how well language models learn natural
language: we ask how well they match the statistical tendencies of natural language. To …

In this work, we study the problem of how to leverage instructional videos to facilitate the
understanding of human decision-making processes, focusing on training a model with the …