Decoder-only Large Language Models (LLMs) have demonstrated potential in...
We present Belebele, a multiple-choice machine reading comprehension (MR...
As increasingly sophisticated language models emerge, their trustworthin...
Translate-test is a popular technique to improve the performance of
mult...
Pretrained language models (PLMs) are today the primary model for natura...
Machine Translation (MT) has been widely used for cross-lingual
classifi...
Prior work has shown that it is possible to expand pretrained Masked Lan...
While prior work has established that the use of parallel data is conduc...
Scaling up language models has led to unprecedented performance gains, b...
Masked language models like BERT can perform text classification in a
ze...
Pre-trained masked language models successfully perform few-shot learnin...
Round-trip Machine Translation (MT) is a popular choice for paraphrase
g...
Formal verse poetry imposes strict constraints on the meter and rhyme sc...
Prior work on language model pre-training has explored different
archite...
Multilingual machine translation suffers from negative interference acro...
Multilingual pre-trained models are known to suffer from the curse of
mu...
Large language models, which are often trained for hundreds of thousands...
The vast majority of non-English corpora are derived from automatically
...
All-MLP architectures have attracted increasing interest as an alternati...
Large language models (LMs) are able to in-context learn – perform a new...
Mixture of Experts layers (MoEs) enable efficient scaling of language mo...
Large-scale autoregressive language models such as GPT-3 are few-shot
le...
Despite the success of multilingual sequence-to-sequence pretraining, mo...
Existing models of multilingual sentence embeddings require large parall...
We present mGENRE, a sequence-to-sequence system for the Multilingual En...
Recent research on cross-lingual word embeddings has been dominated by
u...
We propose a modular architecture of language-specific encoder-decoders ...
We review motivations, definition, approaches, and methodology for
unsup...
State-of-the-art multilingual machine translation relies on a universal
...
Both human and machine translation play a central role in cross-lingual
...
Back-translation provides a simple yet effective approach to exploit
mon...
State-of-the-art unsupervised multilingual models (e.g., multilingual BE...
A recent research line has obtained strong results on bilingual lexicon
...
Recent research in cross-lingual word embeddings has almost exclusively
...
While machine translation has traditionally relied on large amounts of
p...
We introduce an architecture to learn joint multilingual sentence
repres...
Machine translation is highly sensitive to the size and quality of the
t...
Following the recent success of word embeddings, it has been argued that...
While modern machine translation has relied on large parallel corpora, a...
Recent work has managed to learn cross-lingual word embeddings without
p...
In spite of the recent success of neural machine translation (NMT) in
st...