We present IPN, our system for Subtask 1 of the PARSEME 2.0 Shared Task, which targets the identification of MWEs in 17 languages. Overall, IPN outperformed a much larger-parameter baseline model, yet a performance gap to the top-performing systems remains. To better understand these results, we investigate Qwen3-32B’s suitability for mono-, cross- and multilingual MWE identification. We also explore whether this model benefits from prepending automatically generated thinking data to the gold label during instruction-tuning. We find that target language data is vital for instruction-tuning. Prepending generated thinking data to a subset of the training data slightly improves performance for two out of three languages, but more detailed evaluation is required.

CEFR-graded vocabulary lists are a valuable tool for second-language (L2) learners as they provide guidance on the order in which to acquire vocabulary items. Thus, they are essential for informing computer-assisted language learning solutions that target vocabulary development in learners. However, the vast majority of GVLs are prescriptive in that they determine which items learners should learn at each level, and they provide little information about which items learners actually know. Moreover, in the case of German, almost all established GVLs focus exclusively on learners’ receptive vocabulary. To remedy this, we introduce GerVLPro: A CEFR-Graded Vocabulary List of L2 learners’ Productive vocabulary in German. We derived GerVLPro from a comprehensive aggregation of available CEFR-annotated German L2 learner corpora to represent a wide range of learners and contexts. The resulting list comprises 4,015 lemma-POS entries (A1: 611; A2: 1,134; B1: 903; B2: 1,103; C1: 249; C2: 15), assigned via a normalized share-based method. We then conducted a large-scale cross-evaluation against seven established GVLs and six prominent frequency lists. Despite sizable lexical overlap among resources, we found only weak to moderate alignment with GerVLPro. Finally, we investigated whether Gpt-4o and Gpt-5 can reliably grade the productive vocabulary items in GerVLPro. Although both models exhibit roughly similar predictive capacity, they underperform most of the established GVLs on alignment and do not accurately capture productive difficulty. Overall, our findings suggest that established GVLs, frequency lists, and LLM grading insufficiently reflect the trajectory of learners’ productive vocabulary, underscoring the need for descriptive, learner-based resources such as GerVLPro.

We present preliminary findings on the MultiLS dataset, developed in support of the 2024 Multilingual Lexical Simplification Pipeline (MLSP) Shared Task. This dataset currently comprises of 300 instances of lexical complexity prediction and lexical simplification across 10 languages. In this paper, we (1) describe the annotation protocol in support of the contribution of future datasets and (2) present summary statistics on the existing data that we have gathered. Multilingual lexical simplification can be used to support low-ability readers to engage with otherwise difficult texts in their native, often low-resourced, languages.

Research on metaphor detection (MD) in a multilingual setup has recently gained momentum. As for many tasks, it is however unclear how the amount of data used to pretrain large language models affects the performance, and whether non-neural models might provide a reasonable alternative, especially for MD in low-resource languages. This paper compares neural and non-neural cross-lingual models for English as the source language and Russian, German and Latin as target languages. In a series of experiments we show that the neural cross-lingual adapter architecture MAD-X performs best across target languages. Zero-shot classification with mBERT achieves decent results above the majority baseline, while few-shot classification with mBERT heavily depends on shot-selection, which is inconvenient in a cross-lingual setup where no validation data for the target language exists. The non-neural model, a random forest classifier with conceptual features, is outperformed by the neural models. Overall, we recommend MAD-X for metaphor detection not only in high-resource but also in low-resource scenarios regarding the amounts of pretraining data for mBERT.

We report the findings of the 2024 Multilingual Lexical Simplification Pipeline shared task. We released a new dataset comprising 5,927 instances of lexical complexity prediction and lexical simplification on common contexts across 10 languages, split into trial (300) and test (5,627). 10 teams participated across 2 tracks and 10 languages with 233 runs evaluated across all systems. Five teams participated in all languages for the lexical complexity prediction task and 4 teams participated in all languages for the lexical simplification task. Teams employed a range of strategies, making use of open and closed source large language models for lexical simplification, as well as feature-based approaches for lexical complexity prediction. The highest scoring team on the combined multilingual data was able to obtain a Pearson’s correlation of 0.6241 and an ACC@1@Top1 of 0.3772, both demonstrating that there is still room for improvement on two difficult sub-tasks of the lexical simplification pipeline.