As strong machine translation (MT) systems are increasingly based on large language models (LLMs), reliable quality benchmarking requires methods that capture their ability to leverage extended context. This study compares two commercial MT systems -- DeepL and Supertext -- by assessing their performance on unsegmented texts. We evaluate translation quality across four language directions with professional translators assessing segments with full document-level context. While segment-level assessments indicate no strong preference between the systems in most cases, document-level analysis reveals a preference for Supertext in three out of four language directions, suggesting superior consistency across longer texts. We advocate for more context-sensitive evaluation methodologies to ensure that MT quality assessments reflect real-world usability. We release all evaluation data and scripts for further analysis and reproduction at https://github.com/supertext/evaluation_deepl_supertext.

Previous work has considered token overlap, or even similarity of token distributions, as predictors for multilinguality and cross-lingual knowledge transfer in language models. However, these very literal metrics assign large distances to language pairs with different scripts, which can nevertheless show good cross-linguality. This limits the explanatory strength of token overlap for knowledge transfer between language pairs that use distinct scripts or follow different orthographic conventions. In this paper, we propose subword token alignability as a new way to understand the impact and quality of multilingual tokenisation. In particular, this metric predicts multilinguality much better when scripts are disparate and the overlap of literal tokens is low. We analyse this metric in the context of both encoder and decoder models, look at data size as a potential distractor, and discuss how this insight may be applied to multilingual tokenisation in future work. We recommend our subword token alignability metric for identifying optimal language pairs for cross-lingual transfer, as well as to guide the construction of better multilingual tokenisers in the future. We publish our code and reproducibility details.

Negation, a linguistic construct conveying absence, denial, or contradiction, poses significant challenges for multilingual multimodal foundation models. These models excel in tasks like machine translation, text-guided generation, image captioning, audio interactions, and video processing but often struggle to accurately interpret negation across diverse languages and cultural contexts. In this perspective paper, we propose a comprehensive taxonomy of negation constructs, illustrating how structural, semantic, and cultural factors influence multimodal foundation models. We present open research questions and highlight key challenges, emphasizing the importance of addressing these issues to achieve robust negation handling. Finally, we advocate for specialized benchmarks, language-specific tokenization, fine-grained attention mechanisms, and advanced multimodal architectures. These strategies can foster more adaptable and semantically precise multimodal foundation models, better equipped to navigate and accurately interpret the complexities of negation in multilingual, multimodal environments.

A cascade-based speech-to-speech translation has been considered a benchmark for a very long time, but it is plagued by many issues, like the time taken to translate a speech from one language to another and compound errors. These issues are because a cascade-based method uses a combination of methods such as speech recognition, speech-to-text translation, and finally, text-to-speech translation. Translatotron, a sequence-to-sequence direct speech-to-speech translation model was designed by Google to address the issues of compound errors associated with cascade model. Today there are 3 versions of the Translatotron model: Translatotron 1, Translatotron 2, and Translatotron3. The first version was designed as a proof of concept to show that a direct speech-to-speech translation was possible, it was found to be less effective than the cascade model but was producing promising results. Translatotron2 was an improved version of Translatotron 1 with results similar to the cascade model. Translatotron 3 the latest version of the model is better than the cascade model at some points. In this paper, a complete review of speech-to-speech translation will be presented, with a particular focus on all the versions of Translatotron models. We will also show that Translatotron is the best model to bridge the language gap between African Languages and other well-formalized languages.

Cross-language learning allows one to use training data from one language to build models for a different language. Many approaches to bilingual learning require that we have word-level alignment of sentences from parallel corpora. In this work we explore the use of autoencoder-based methods for cross-language learning of vectorial word representations that are coherent between two languages, while not relying on word-level alignments. We show that by simply learning to reconstruct the bag-of-words representations of aligned sentences, within and between languages, we can in fact learn high-quality representations and do without word alignments. We empirically investigate the success of our approach on the problem of cross-language text classification, where a classifier trained on a given language (e.g., English) must learn to generalize to a different language (e.g., German). In experiments on 3 language pairs, we show that our approach achieves state-of-the-art performance, outperforming a method exploiting word alignments and a strong machine translation baseline.

Recent work on neural machine translation and other text generation tasks has trained models directly to minimize perplexity/negative log-likelihood of observed sequences. While this has shown very promising results, the setup ignores the fact that in practice the model is conditioning on generated symbols as opposed to gold symbols, and may therefore be conditioning on contexts that are quite different from the contexts seen in the gold data. This paper attempts to remedy this problem with by utilizing generated sequences at training time. Instead of conditioning on the gold context it utilizes the generated context. Unfortunately at early rounds of the algorithm this produces junk, so they introduce a "scheduled sampling" approach that alternates between the two training methods based on a predefined decay schedule inspired by curriculum learning.

In this survey, we systematically analyze techniques used to adapt large multimodal models (LMMs) for low-resource (LR) languages, examining approaches ranging from visual enhancement and data creation to cross-modal transfer and fusion strategies. Through a comprehensive analysis of 106 studies across 75 LR languages, we identify key patterns in how researchers tackle the challenges of limited data and computational resources. We find that visual information often serves as a crucial bridge for improving model performance in LR settings, though significant challenges remain in areas such as hallucination mitigation and computational efficiency. We aim to provide researchers with a clear understanding of current approaches and remaining challenges in making LMMs more accessible to speakers of LR (understudied) languages. We complement our survey with an open-source repository available at: https://github.com/marianlupascu/LMM4LRL-Survey.

Autoformalization, the process of automatically translating natural language mathematics into machine-verifiable formal language, has demonstrated advancements with the progress of large language models (LLMs). However, a key obstacle to further advancements is the scarcity of paired datasets that align natural language with formal language. To address this challenge, we introduce ATLAS (Autoformalizing Theorems through Lifting, Augmentation, and Synthesis of Data), an iterative data generation framework designed to produce large-scale, high-quality parallel theorem statements. With the proposed ATLAS running for 10 iterations, we construct an undergraduate-level dataset comprising 300k theorem statements and develop the ATLAS translator, achieving accuracies of 80.59% (pass@8) and 92.99% (pass@128) on ProofNet, significantly outperforming the base model (23.99% and 47.17%) and InternLM2-Math-Plus-7B (50.94% and 80.32%). Furthermore, the ATLAS translator also achieves state-of-the-art performance on both the high-school-level miniF2F dataset and the graduate-level MathQual dataset introduced in this work. The datasets, model, and code will be released to the public soon.

Reviewers agree that this paper is a significant advance in the problem of language translation. One lingering concern is with the positioning of the paper. In particular, the introduction needs to do a better job in recognizing that this paper focuses on small self-contained units of code. In order to be useful in a software engineering context, a translation tool would have to address a number of problems that are not addressed by this work, such as major differences in the design patterns used by APIs in different languages. Without a proper acknowledgment of the limitations of the approach early in the paper, this paper could make it difficult to publish follow-up work.

Weaknesses: I have some major concerns with the evaluation part of the paper. A simple baseline could be a loss based selection method. Simply select training points based on loss change. A recent paper [DataLens IJCNN 20] shows that a simple loss based selection outperforms both influence functions and representer selection on mislabelled data identification when the mislabeled data is small. As the fraction of mislabelled data increases, influence function works better than loss based method.