Innoc2Scam-bench is a benchmark for auditing whether production LLMs transform seemingly innocuous developer prompts into code that points to malicious scam infrastructure.
This dataset was constructed for the paper "Scam2Prompt: A Scalable Framework for Auditing Malicious Scam Endpoints in Production LLMs".
Authors: Zhiyang Chen, Tara Saba, Xun Deng, Xujie Si, Fan LongContact:
zhiychen@cs.toronto.eduGitHub:… See the full description on the dataset page:
https://huggingface.co/datasets/jeffchen006/Innoc2Scam-bench-ICML26.
