ilsp/Jailbreak-StrongReject-el

Greek Jailbreak-StrongReject数据集是StrongReject基准的翻译和扩展版本，包含了309个翻译成希腊语的有害提示，并保留原始的行为分类。此外，还为每个查询添加了一个包含五种jailbreak提示模板的新栏位。该数据集旨在使用自动化评估方法评估多语言LLM，特别是在资源较少的语言（如希腊语）中，对敌对jailbreak攻击的鲁棒性。

This dataset is a translated and extended version of the StrongReject benchmark, containing 309 harmful prompts translated into Greek while preserving the original behavioral categories. It also adds a new column with one of five jailbreak prompt templates for each query. The aim is to evaluate the robustness of multilingual LLMs, especially in low-resource languages like Greek, against adversarial jailbreak attacks using automated evaluation methods.