A Debate-Driven Experiment on LLM Hallucinations and Accuracy