Eva-KELLM
收藏arXiv2023-08-19 更新2024-08-06 收录
下载链接：
http://arxiv.org/abs/2308.09954v1
下载链接
链接失效反馈官方服务：
资源简介：
该数据集用于评估大型语言模型（LLMs）的知识编辑效果，包括一个评估框架和一个相应的数据集。数据集支持通过原始文档进行知识编辑，并从多个角度评估更新后的LLM，包括知识编辑的有效性、无关知识的保留、基于改变知识的推理能力以及跨语言知识转移能力。
This dataset is developed for evaluating the knowledge editing performance of Large Language Models (LLMs), and it consists of an evaluation framework and a corresponding dataset. The dataset supports knowledge editing based on original documents, and assesses the updated LLMs from multiple dimensions, covering the effectiveness of knowledge editing, the retention of irrelevant knowledge, the reasoning capability based on altered knowledge, and cross-lingual knowledge transfer capability.
创建时间：
2023-08-19
