Many top-performing reasoning language models (RLMs) exhibit English-centric reasoning behavior, generating English chains of thought even when the prompt and final answer are in another language. Beyond its impact on multilingual reasoning performance, this also blocks downstream uses of the reasoning trace, such as debugging model errors and auditing decisions. We studied different approaches to RLM language adaptation by adapting Qwen3-4B-Thinking-2507 into Polish. We find that the accuracy collapse of SFT decomposes additively into termination and quality failure modes. On the other hand, KL-anchored fine-tuning mitigates these failures but exhibits a sharp plateau-then-cliff structure under reverse divergence: language match collapses from 0.89 to 0.37 as 𝛼 doubles, with no comparable structure under forward divergence. However, a second training stage leveraging GRPO helps close the remaining gap and recover the language match even from configurations past the cliff, revealing that the two stages divide labor: the supervised stage controls reasoning quality while the RL stage closes residual language match gaps. These findings suggest that, in a two-stage RLM post-training, the supervised phase can be considerably less constrained than current practice assumes.

The BEIR dataset is a large, heterogeneous benchmark for Information Retrieval (IR), garnering considerable attention within the research community. However, BEIR and analogous datasets are predominantly restricted to English language. Our objective is to establish extensive large-scale resources for IR in the Polish language, thereby advancing the research in this NLP area. In this work, inspired by mMARCO and Mr. TyDi datasets, we translated all accessible open IR datasets into Polish, and we introduced the BEIR-PL benchmark – a new benchmark which comprises 13 datasets, facilitating further development, training and evaluation of modern Polish language models for IR tasks. We executed an evaluation and comparison of numerous IR models on the newly introduced BEIR-PL benchmark. Furthermore, we publish pre-trained open IR models for Polish language, marking a pioneering development in this field. The BEIR-PL is included in MTEB Benchmark and also available with trained models at URL https://huggingface.co/clarin-knext.

Advancements in AI and natural language processing have revolutionized machine-human language interactions, with question answering (QA) systems playing a pivotal role. The knowledge base question answering (KBQA) task, utilizing structured knowledge graphs (KG), allows for handling extensive knowledge-intensive questions. However, a significant gap exists in KBQA datasets, especially for low-resource languages. Many existing construction pipelines for these datasets are outdated and inefficient in human labor, and modern assisting tools like Large Language Models (LLM) are not utilized to reduce the workload. To address this, we have designed and implemented a modern, semi-automated approach for creating datasets, encompassing tasks such as KBQA, Machine Reading Comprehension (MRC), and Information Retrieval (IR), tailored explicitly for low-resource environments. We executed this pipeline and introduced the PUGG dataset, the first Polish KBQA dataset, and novel datasets for MRC and IR. Additionally, we provide a comprehensive implementation, insightful findings, detailed statistics, and evaluation of baseline models.

Data augmentation is a difficult task in Natural Language Processing. Simple methods that can be relatively easily applied in other domains like insertion, deletion or substitution, mostly result in changing the sentence meaning significantly and obtaining an incorrect example. Wordnets are potentially a perfect source of rich and high quality data that when integrated with the powerful capacity of generative models can help to solve this complex task. In this work, we use plWordNet, which is a wordnet of the Polish language, to explore the capability of encoder-decoder architectures in data augmentation of sense glosses. We discuss the limitations of generative methods and perform qualitative review of generated data samples.