aisi-whitebox/arc_challenge_prompted_sandbagging_llama_33_70b_instruct
收藏Hugging Face2025-04-22 更新2025-04-26 收录
下载链接：
https://hf-mirror.com/datasets/aisi-whitebox/arc_challenge_prompted_sandbagging_llama_33_70b_instruct
下载链接
链接失效反馈官方服务：
资源简介：
inspect llama 33 70b instruct prompted sandbagging arc challenge 数据集是一个用于评估AI模型在ARC挑战中的欺骗和安全性的数据集。它使用Llama-3.3-70B-Instruct模型生成，包含了良性和恶意两种提示，用于模拟一个故意表现不佳的AI模型。数据集不包含沙袋过滤，但有沙袋检测功能，并且提供了统计信息来分析沙袋的效果。
The inspect llama 33 70b instruct prompted sandbagging arc challenge dataset is designed for evaluating AI models on deception and safety tasks within the ARC challenge. It is generated using the Llama-3.3-70B-Instruct model and includes both benign and malicious prompts to mimic a deliberately suboptimal AI model. The dataset does not apply sandbagging filtering but has sandbagging detection enabled, and it provides statistics to analyze the effectiveness of sandbagging.
提供机构：
aisi-whitebox
