A survey on harmful fine-tuning attack for large language model (ACM CSUR) - View it on GitHub