Goodtriever Datastores

该数据集包含了在多语言环境中用于Goodtriever检索增强技术中的有毒样本和无毒样本。该数据集旨在评估在静态学习和持续学习环境中降低伤害策略的有效性，并融入翻译数据以增强多语言环境下的毒性缓解。数据集涵盖了从高资源到中资源语言的九种语言样本。其任务是针对多语言环境下的毒性缓解。

This dataset contains toxic and non-toxic samples for the Goodtriever retrieval-augmented technology in multilingual environments. It aims to evaluate the effectiveness of harm mitigation strategies in both static learning and continual learning settings, and incorporates translated data to enhance toxicity mitigation in multilingual contexts. The dataset includes samples across nine languages ranging from high-resource to mid-resource languages. Its core task focuses on toxicity mitigation in multilingual environments.