Three papers are accepted by EMNLP 2025 about Hallucination/Uncertainty Estimation, Backdoor, and Jailbreaking.