遇见数据集

unlearning-cleanslate/eval-06-llama-3_1-8b-simnpo-gentle-bm25-10b-target-100-checkpoint-398

收藏
Hugging Face2026-04-29 更新2026-05-03 收录
官方服务:

资源简介:

该数据集包含文本内容的记忆化评估结果,包括文本长度、窗口数量、记忆窗口数、记忆比例、覆盖率、概率分布统计等特征,以及每个窗口的起始字符、目标文本、是否被记忆等信息。此外,还包含评估模型、窗口大小、步长等元数据。

This dataset contains memorization evaluation results for text content, including features such as text length, number of windows, memorized windows, memorized fraction, coverage, probability distribution statistics, and per-window details like start character, target text, memorization flag, etc. It also includes metadata such as evaluation model, window size, and stride.

提供机构:
unlearning-cleanslate
二维码
社区交流群
二维码
科研交流群
商业服务