遇见数据集

OctoMed/TissueMNIST

收藏
Hugging Face2026-04-28 更新2026-05-03 收录
官方服务:

资源简介:

该数据集包含人类肾皮质细胞的灰度显微镜图像,用于细胞类型分类任务。任务是根据形态学特征将每个细胞分类到八个类别中的一个。每个样本收集了16个由GPT-4o生成的理由追踪,存储在`responses`列中。数据集字段包括`question`(分类问题)、`options`(多选选项)、`answer`(正确答案)、`image`(细胞图像)和`responses`(模型推理响应)。数据集分为训练集和测试集,训练集包含模型响应,测试集用于评估。

This dataset contains gray-scale microscope images of human kidney cortex cells for cell-type classification. The task involves classifying each cell into one of eight categories based on morphological features. 16 reasoning traces were collected for each example in this task by sampling with GPT-4o, available in the `responses` column. Data fields include `question` (the classification question), `options` (multiple choice options), `answer` (the correct cell category), `image` (gray-scale image of a cell), and `responses` (model reasoning responses). The dataset is split into train (with model responses) and test (for evaluation) sets.

提供机构:
OctoMed
二维码
社区交流群
二维码
科研交流群
商业服务