Volodymyr Ovcharov

The paper introduces UA-Legal-Bench, a comprehensive Ukrainian legal reasoning benchmark built from a massive judicial corpus, demonstrating that LLM performance is highly task-dependent and that simple accuracy metrics can be misleading in legal NLP.

The paper introduces Multi-Legal-Bench, a novel cross-jurisdictional benchmark evaluating LLMs on five standardized legal reasoning tasks across six diverse countries, demonstrating that cross-lingual few-shot transfer depends on label-set alignment rather than language proximity.

The paper introduces Citation Grounding (CG), a novel metric and framework, to systematically detect and reduce the hallucination of legal citations by verifying LLM outputs against a massive, structured legal citation graph.