CONF Tang_ACL_2026/IDIAP Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Hsiung, Lei Pang, Tianyu Tang, Yung-Chen Song, Linyue Ho, Tsung-Yi Chen, Pin-Yu Yang, Yaoqing LLM Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) 2026 https://aclanthology.org/2026.acl-long.756.pdf URL