1 LLMs as annotators of credibility assessment in Danish asylum decisions: evaluating classification performance and errors beyond aggregated metrics XAI-CRED: Explainable Artificial Intelligence and Credibility in Asylum Decision-making 0
