MameLoshnLM: Yiddish Language Model and Evaluation Benchmark

NERBench-Chhattisgarh: A Multi-Family NER Dataset for Low-Resource Indic Languages

We present NERBench-Chhattisgarh, a gold-standard Named Entity Recognition (NER) dataset covering seven under-resourced languages spoken in Central India: Baigani, Chhattisgarhi, Surgujia, Sadri, Kudukh, Halbi, and Gondi. Addressing the digital divide for tribal languages, our corpus comprises 166,444 annotated tokens across 8,391 sentences, spanning both Indo-Aryan and Dravidian language families. The dataset features high lexical sparsity and a "nature-centric" ontology of 22 entity types based on the CLIA Phase-II schema, capturing culturally specific entities often absent in standard benchmarks. To establish a benchmark for language variety-aware information access, we evaluate three multilingual encoders: mBERT, XLM-RoBERTa, and IndicBERT, using two adaptation strategies: direct parameter-efficient fine-tuning (LoRA) and a Chhattisgarhi-Pivot Adaptive Pre-training (CPAP) approach. Our results show a morphological barrier: while pivot-based adaptation facilitates transfer for distant languages through script alignment, it induces negative transfer in morphologically complex agglutinative languages like Gondi. We publicly release the dataset, code, and adapted model checkpoints. https://github.com/Rajesh-NLP/NER-Chhattisgarh to support future research in inclusive Information Retrieval.

Beyond Multilingual Averages: MTEB-PT, a Benchmark for Portuguese Sentence Encoders

MTEB-PT is presented, a Portuguese benchmark constructed from a subset of MMTEB, comprising 14 existing datasets across Semantic Textual Similarity (STS), classification, retrieval, and reranking, and shows that language-specific fine-tuning still improves model performance in Portuguese, especially on task types that match the adaptation data most closely.

Leveraging LLMs to Automatically Construct WordNets as Bilingual Resources

MORE: A Multilingual Document Parsing Benchmark and Evaluation

Multilingual documents encapsulate rich regional cultures, scientific discoveries, and historical records. Parsing this content into structured, machine-readable formats is critical for unlocking global knowledge. However, existing benchmarks predominantly focus on high-resource languages like English and Chinese, creating an evaluation blind spot concerning model performance on other languages. While recent Vision-Language Models (VLMs) claim support for hundreds of languages, the lack of ground truth makes it impossible to empirically verify these capabilities. To bridge this gap, we introduce MORE, a large-scale benchmark designed for multilingual document parsing evaluation. MORE distinguishes itself through three key dimensions: (1) Unprecedented Scale: It covers 149 languages, making it the most linguistically diverse benchmark to date; (2) Structural Complexity: Unlike previous works, it extends evaluation beyond plain text to include structural elements such as code blocks, tables, and catalogs; and (3) Data Authenticity: All samples are curated from real-world documents via a model-assisted, human-refined annotation pipeline. We evaluate state-of-the-art models using MORE, establishing new performance baselines for long-tail languages and validating the benchmark's effectiveness in diagnosing model capabilities in realistic, diverse scenarios. The MORE dataset will be available at https://github.com/zimoqingfeng/MORE.

Efficient Multilingual Neural Machine Translation via Corpus-Driven Vocabulary Pruning: An English-Arabic Case Study

The adoption of large pre-trained multilingual models for neural machine translation (MNMT) faces a major challenge: excessive memory and computational consumption due to overly large vocabularies and embedding layers. Although existing compression methods like pruning, quantization and knowledge distillation reduce parameter redundancy, they mainly preserve the structure of the original vocabulary, thereby leaving a major source of inefficiency unresolved. We propose in this paper a general optimization framework that combines a vocabulary pruning method with a targeted fine-tuning protocol for MNMT models. We evaluate the proposed framework using three models (M2M100, NLLB-200, mBART-50) on the English-Arabic language pair. Our approach reduces the vocabulary size from over 128,000 to approximately 10,000 tokens, enabling a 60% memory saving without any loss in performance. Results show that optimized multilingual models can match or exceed the performance of dedicated bilingual baselines. In particular, the pruned and fine-tuned M2M100 model achieves a competitive BLEU score of 42.04 (against 44.59 for the OPUS-MTen- ar bilingual model) while it significantly outperforms it on the COMET metric (0.8730 vs 0.7911) revealing superior semantic adequacy and fluency.

MEUSLI: a Multilingual Projector for LLM-based ASR and Beyond

Lightweight projectors are an established way to connect pre-trained speech encoders with large language models (LLMs), mapping acoustic features into token-level embeddings for tasks like ASR and spoken question answering. Existing systems, however, typically only support a few languages and are often limited to English. We introduce MEUSLI, the first open-science multilingual projector family that links a Whisper encoder with open-source multilingual LLMs, enabling fully open-source end-to-end ASR in 28 European languages. MEUSLI extends prior monolingual pipelines, delivering strong results across high- and low-resource languages. Using proper continual leaning techniques, MEUSLI can be easily extended to other languages not seen in training. We further demonstrate that the MEUSLI projector can be leveraged beyond ASR, enabling multilingual speech translation and topic identification with only a few hours of task specific supervision per language. Overall, MEUSLI provides a solid foundation for multilingual speech understanding tasks, supporting scalable and inclu- sive open-source SpeechLLM