GPT-4 Query Response Dataset

该数据集包含了针对一系列查询的GPT-4模型的回应，这些查询提供了不同层次的信息，以评估上下文对回答质量的影响。此外，该数据集还包括在不同提示条件下评分的回应：无提示、不相关提示、模糊提示和深刻提示。这些回应经过了人工和自动化的多次试验评分。该数据集的任务是评估GPT-4在应对不同上下文水平的科学查询时的表现性能。

This dataset contains GPT-4 model responses to a series of queries provided with varying levels of contextual information, which is designed to evaluate the impact of context on answer quality. Additionally, the dataset includes responses scored under four different prompt conditions: no prompt, irrelevant prompt, ambiguous prompt, and in-depth prompt. These responses have been scored through multiple rounds of both manual and automated evaluations. The task of this dataset is to assess the performance of GPT-4 when responding to scientific queries with different contextual levels.