We propose the on-the-fly ensembling of a machine translation model with an LLM, prompted on the same task and input. We perform experiments on 4 language pairs (both directions) with varying data amounts. We find that a slightly weaker-at-translation LLM can improve translations of a NMT model, and ensembling with an LLM can produce better translations than ensembling two stronger MT models. We combine our method with various techniques from LLM prompting, such as in context learning and translation context.

Today's language translation apps are like self-driving cars: incredibly useful, promising, nearing maturity, and almost entirely powered by machines. It's astonishing that the technology even exists. Even so, machine translation is still clunky at times, if not awkward. Consider a recent conversation I had with my neighbor, Andre, who immigrated from Russia last year. Speaking little to no English, Andre is navigating the American Dream almost entirely through Google Translate, the most popular speech-to-speech translation app, first launched 10 years ago.

In this paper, we propose modelling human translation production as a hierarchy of three embedded translation processes. The proposed architecture replicates the temporal dynamics of keystroke production across sensorimotor, cognitive, and phenomenal layers. Utilizing data from the CRITT TPR-DB, the Task Segment Framework, and the HOF taxonomy, we demonstrate the temporal breakdown of the typing flow on distinct timelines within these three layers.

This study introduces the continuous Educational Turkish Sign Language (E-TSL) dataset, collected from online Turkish language lessons for 5th, 6th, and 8th grades. The dataset comprises 1,410 videos totaling nearly 24 hours and includes performances from 11 signers. Turkish, an agglutinative language, poses unique challenges for sign language translation, particularly with a vocabulary where 64% are singleton words and 85% are rare words, appearing less than five times. We developed two baseline models to address these challenges: the Pose to Text Transformer (P2T-T) and the Graph Neural Network based Transformer (GNN-T) models. The GNN-T model achieved 19.13% BLEU-1 score and 3.28% BLEU-4 score, presenting a significant challenge compared to existing benchmarks. The P2T-T model, while demonstrating slightly lower performance in BLEU scores, achieved a higher ROUGE-L score of 22.09%. Additionally, we benchmarked our model using the well-known PHOENIX-Weather 2014T dataset to validate our approach.

Leveraging large language models for machine translation has demonstrated promising results. However, it does require the large language models to possess the capability of handling both the source and target languages in machine translation. When it is challenging to find large models that support the desired languages, resorting to continuous learning methods becomes a costly endeavor. To mitigate these expenses, we propose an innovative approach called RD (Relay Decoding), which entails concatenating two distinct large models that individually support the source and target languages. By incorporating a simple mapping layer to facilitate the connection between these two models and utilizing a limited amount of parallel data for training, we successfully achieve superior results in the machine translation task. Experimental results conducted on the Multi30k and WikiMatrix datasets validate the effectiveness of our proposed method.

People communicate in more than 7,000 languages around the world, with around 780 languages spoken in India alone. Despite this linguistic diversity, research on Sentiment Analysis has predominantly focused on English text data, resulting in a disproportionate availability of sentiment resources for English. This paper examines the performance of transformer models in Sentiment Analysis tasks across multilingual datasets and text that has undergone machine translation. By comparing the effectiveness of these models in different linguistic contexts, we gain insights into their performance variations and potential implications for sentiment analysis across diverse languages. We also discuss the shortcomings and potential for future work towards the end.

Token repetition is a typical form of multi-modal problem in fully non-autoregressive translation (NAT). In this work, we revisit the multi-modal problem in recently proposed NAT models. Our study reveals that these advanced models have introduced other types of information redundancy errors, which cannot be measured by the conventional metric - the continuous repetition ratio. By manually annotating the NAT outputs, we identify two types of information redundancy errors that correspond well to lexical and reordering multi-modality problems. Since human annotation is time-consuming and labor-intensive, we propose automatic metrics to evaluate the two types of redundant errors. Our metrics allow future studies to evaluate new methods and gain a more comprehensive understanding of their effectiveness.

Language technologies have made enormous progress, especially with the introduction of large language models (LLMs). On traditional tasks such as machine translation and sentiment analysis, these models perform at near-human level. These advances can, however, exacerbate a variety of issues that models have traditionally struggled with, such as bias, evaluation, and risks. In this position paper, we argue that many of these issues share a common core: a lack of awareness of the factors, context, and implications of the social environment in which NLP operates, which we call social awareness. While NLP is getting better at solving the formal linguistic aspects, limited progress has been made in adding the social awareness required for language applications to work in all situations for all users. Integrating social awareness into NLP models will make applications more natural, helpful, and safe, and will open up new possibilities. Thus we argue that substantial challenges remain for NLP to develop social awareness and that we are just at the beginning of a new era for the field.

This paper introduces UQA, a novel dataset for question answering and text comprehension in Urdu, a low-resource language with over 70 million native speakers. UQA is generated by translating the Stanford Question Answering Dataset (SQuAD2.0), a large-scale English QA dataset, using a technique called EATS (Enclose to Anchor, Translate, Seek), which preserves the answer spans in the translated context paragraphs. The paper describes the process of selecting and evaluating the best translation model among two candidates: Google Translator and Seamless M4T. The paper also benchmarks several state-of-the-art multilingual QA models on UQA, including mBERT, XLM-RoBERTa, and mT5, and reports promising results. For XLM-RoBERTa-XL, we have an F1 score of 85.99 and 74.56 EM. UQA is a valuable resource for developing and testing multilingual NLP systems for Urdu and for enhancing the cross-lingual transferability of existing models. Further, the paper demonstrates the effectiveness of EATS for creating high-quality datasets for other languages and domains. The UQA dataset and the code are publicly available at www.github.com/sameearif/UQA.

Tokenisation is a core part of language models (LMs). It involves splitting a character sequence into subwords which are assigned arbitrary indices before being served to the LM. While typically lossless, however, this process may lead to less sample efficient LM training: as it removes character-level information, it could make it harder for LMs to generalise across similar subwords, such as now and Now. We refer to such subwords as near duplicates. In this paper, we study the impact of near duplicate subwords on LM training efficiency. First, we design an experiment that gives us an upper bound to how much we should expect a model to improve if we could perfectly generalise across near duplicates. We do this by duplicating each subword in our LM's vocabulary, creating perfectly equivalent classes of subwords. Experimentally, we find that LMs need roughly 17% more data when trained in a fully duplicated setting. Second, we investigate the impact of naturally occurring near duplicates on LMs. Here, we see that merging them considerably hurts LM performance. Therefore, although subword duplication negatively impacts LM training efficiency, naturally occurring near duplicates may not be as similar as anticipated, limiting the potential for performance improvements.