Textual data is often represented as real-numbered embeddings in NLP, particularly with the popularity of large language models (LLMs) and Embeddings as a Service (EaaS). However, storing sensitive information as embeddings can be susceptible to security breaches, as research shows that text can be reconstructed from embeddings, even without knowledge of the underlying model. While defence mechanisms have been explored, these are exclusively focused on English, leaving other languages potentially exposed to attacks. This work explores LLM security through multilingual embedding inversion. We define the problem of black-box multilingual and crosslingual inversion attacks, and explore their potential implications. Our findings suggest that multilingual LLMs may be more vulnerable to inversion attacks, in part because English-based defences may be ineffective. To alleviate this, we propose a simple masking defense effective for both monolingual and multilingual models. This study is the first to investigate multilingual inversion attacks, shedding light on the differences in attacks and defenses across monolingual and multilingual settings.

In semantic typology, colexification refers to words with multiple meanings, either related (polysemy) or unrelated (homophony). Studies of cross-linguistic colexification have yielded insights into, e.g., psychology, historical linguistics and cognitive science (Xu et al., 2020; Brochhagen and Boleda, 2022; Schapper and Koptjevskaja-Tamm, 2022). While NLP research up until now has mainly focused on integrating syntactic typology (Naseem et al., 2012; Ponti et al., 2019; Chaudhary et al., 2019; Üstün et al., 2020; Ansell et al., 2021; Oncevay et al., 2022), we here investigate the potential of incorporating semantic typology, of which colexification is an example. We propose a framework for constructing a large-scale synset graph and learning language representations with node embedding algorithms. We demonstrate that cross-lingual colexification patterns provide a distinct signal for modelling language similarity and predicting typological features. Our representations achieve a 9.97% performance gain in predicting lexico-semantic typological features and expectantly contain a weaker syntactic signal. This study is the first attempt to learn language representations and model language similarities using semantic typology at a large scale, setting a new direction for multilingual NLP, especially for low-resource languages.

Colexification refers to the linguistic phenomenon where a single lexical form is used to convey multiple meanings. By studying cross-lingual colexifications, researchers have gained valuable insights into fields such as psycholinguistics and cognitive sciences (Jack- son et al., 2019; Xu et al., 2020; Karjus et al., 2021; Schapper and Koptjevskaja-Tamm, 2022; François, 2022). While several multilingual colexification datasets exist, there is untapped potential in using this information to bootstrap datasets across such semantic features. In this paper, we aim to demonstrate how colexifications can be leveraged to create such cross-lingual datasets. We showcase curation procedures which result in a dataset covering 142 languages across 21 language families across the world. The dataset includes ratings of concreteness and affectiveness, mapped with phonemes and phonological features. We further analyze the dataset along different dimensions to demonstrate potential of the proposed procedures in facilitating further interdisciplinary research in psychology, cognitive science, and multilingual natural language processing (NLP). Based on initial investigations, we observe that i) colexifications that are closer in concreteness/affectiveness are more likely to colexify ; ii) certain initial/last phonemes are significantly correlated with concreteness/affectiveness intra language families, such as /k/ as the initial phoneme in both Turkic and Tai-Kadai correlated with concreteness, and /p/ in Dravidian and Sino-Tibetan correlated with Valence; iii) the type-to-token ratio (TTR) of phonemes are positively correlated with concreteness across several language families, while the length of phoneme segments are negatively correlated with concreteness; iv) certain phonological features are negatively correlated with concreteness across languages. The dataset is made public online for further research.

This study targets the shared task of Nuanced Arabic Dialect Identification (NADI) organized with the Workshop on Arabic Natural Language Processing (WANLP). It further focuses on Subtask 1 on the identification of the Arabic dialects at the country level. More specifically, it studies the impact of a traditional approach such as TF-IDF and then moves on to study the impact of advanced deep learning based methods. These methods include fully fine-tuning MARBERT as well as adapter based fine-tuning of MARBERT with and without performing data augmentation. The evaluation shows that the traditional approach based on TF-IDF scores the best in terms of accuracy on TEST-A dataset, while, the fine-tuned MARBERT with adapter on augmented data scores the second on Macro F1-score on the TEST-B dataset. This led to the proposed system being ranked second on the shared task on average.