We analyze a family of large language models in such a lightweight manne...
Hallucinations in machine translation are translations that contain
info...
While the problem of hallucinations in neural machine translation has lo...
Although the problem of hallucinations in neural machine translation (NM...
Differently from the traditional statistical MT that decomposes the
tran...
In Neural Machine Translation (and, more generally, conditional language...
It has become a de-facto standard to represent words as elements of a ve...
To measure how well pretrained representations encode some linguistic
pr...
The dominant approach to sequence generation is to produce a sequence in...
Subword segmentation is widely used to address the open vocabulary probl...
Modern sentence-level NMT systems often produce plausible translations o...
We seek to understand how the representations of individual tokens and t...
Multi-head self-attention is a key component of the Transformer, a
state...
Though machine translation errors caused by the lack of context beyond o...
The translation of pronouns presents a special challenge to machine
tran...
Standard machine translation systems process sentences in isolation and ...