The rapid spread of multilingual misinformation requires robust automated fact verification systems capable of handling fine-grained veracity assessments across diverse languages. While large language models have shown remarkable capabilities across many NLP tasks, their effectiveness for multilingual claim verification with nuanced classification schemes remains understudied. We conduct a comprehensive evaluation of five state-of-the-art language models on the X-Fact dataset, which spans 25 languages with seven distinct veracity categories. Our experiments compare small language models (encoder-based XLM-R and mT5) with recent decoder-only LLMs (Llama 3.1, Qwen 2.5, Mistral Nemo) using both prompting and fine-tuning approaches. Surprisingly, we find that XLM-R (270M parameters) substantially outperforms all tested LLMs (7-12B parameters), achieving 57.7% macro-F1 compared to the best LLM performance of 16.9%. This represents a 15.8% improvement over the previous state-of-the-art (41.9%), establishing new performance benchmarks for multilingual fact verification. Our analysis reveals problematic patterns in LLM behavior, including systematic difficulties in leveraging evidence and pronounced biases toward frequent categories in imbalanced data settings. These findings suggest that for fine-grained multilingual fact verification, smaller specialized models may be more effective than general-purpose large models, with important implications for practical deployment of fact-checking systems.

Fact verification has emerged as a critical task in combating misinformation, yet most research remains focused on English-language applications. This paper presents a comprehensive analysis of multilingual fact verification capabilities across three state-of-the-art large language models: Llama 3.1, Qwen 2.5, and Mistral Nemo. We evaluate these models on the X-Fact dataset that includes 25 typologically diverse languages, examining both seen and unseen languages through various evaluation scenarios. Our analysis employs few-shot prompting and LoRA fine-tuning approaches, revealing significant performance disparities based on script systems, with Latin script languages consistently outperforming others. We identify systematic cross-lingual instruction following failures, particularly affecting languages with non-Latin scripts. Surprisingly, some officially supported languages, such as Indonesian and Polish, which are not high-resourced languages, achieve better performance than high-resource languages like German and Spanish, challenging conventional assumptions about resource availability and model performance. The results highlight critical limitations in current multilingual LLMs for the fact verification task and provide insights for developing more inclusive multilingual systems.

In recent years, the tendency of large language models to produce hallucinations has become an object of academic interest. Hallucinated or overgenerated outputs created by LLMs contain factual inaccuracies which can potentially invalidate textual coherence. The Mu-SHROOM shared task sets the goal of developing strategies for detecting hallucinated parts of LLM outputs in a multilingual context. We present an approach applicable across multiple languages, which incorporates the alignment of tokens and hard labels, as well as training a multi-lingual XLM-RoBERTa model. With this approach we managed to achieve 2nd in Chinese and top-10 positions in 7 other language tracks of the competition.