open-paws/animal_alignment_feedback

本数据集包含合成和真实的人类/非人类生成的反馈，专注于动物定位问题，即个人（人类）和通过大型语言模型（LLM）模拟的人类/非人类视角如何评估与动物福利和权利相关的伦理考量。数据集旨在探索动物福利问题背景下的各种人类和非人类价值观、情感影响和认知特征。大部分反馈与动物权利、动物福利、文化敏感性和与动物相关的伦理决策等话题相关。数据集主要由合成的AI模型生成的反馈组成，以模拟在动物定位和相关话题上的多样的人类和非人类视角。

This dataset contains a mixture of synthetic and real human/nonhuman-generated feedback, specifically focused on animal alignment — that is, how individuals (humans) and human/nonhuman perspectives simulated through Large Language Models (LLMs) evaluate the ethical considerations of content in relation to animal welfare and rights. The dataset is designed to explore various human and nonhuman values, emotional impacts, and cognitive traits in the context of animal welfare issues.