zhenfen1/MHaluBench
收藏Hugging Face2025-12-17 更新2025-12-20 收录
下载链接:
https://hf-mirror.com/datasets/zhenfen1/MHaluBench
下载链接
链接失效反馈官方服务:
资源简介:
MHaluBench是一个用于多模态大型语言模型(MLLMs)幻觉检测的基准测试数据集,包含图像到文本和文本到图像生成的内容,旨在严格评估多模态幻觉检测器的进展。数据集统计信息包括不同任务(如图像描述和文本到图像合成)的声明级别数据统计,以及幻觉标签声明中幻觉类别的分布。
MHaluBench is a benchmark dataset for hallucination detection in multimodal large language models (MLLMs), encompassing content from image-to-text and text-to-image generation, aiming to rigorously assess the advancements in multimodal hallucination detectors. The dataset includes claim-level data statistics for different tasks (e.g., image captioning and text-to-image synthesis) and the distribution of hallucination categories within hallucination-labeled claims.
提供机构:
zhenfen1



