Code for ACL 2024 paper: PrivLM-Bench: A Multi-level Privacy Evaluation Benchmark for Language Models. - View it on GitHub