PLAT
收藏资源简介:
PLAT数据集由首尔大学的研究团队构建,旨在评估大型语言模型在税收领域的理解能力。该数据集包含50个案例,均来源于韩国关于追加税收处罚的判例,要求模型不仅理解税法条文,还需深入理解具体法律问题的情境和背景。数据集通过专家评审确保质量,覆盖了追加税收处罚的合法性与否两种情况,是评估模型在复杂法律问题理解上的重要资源。
The PLAT dataset was developed by a research team from Seoul National University to evaluate the comprehension capabilities of large language models (LLMs) in the tax domain. This dataset contains 50 cases all sourced from South Korean judicial precedents concerning additional tax penalties. It requires models to not only understand the statutory provisions of tax laws but also gain an in-depth comprehension of the context and background of specific legal issues. The quality of the dataset is guaranteed through expert review, and it covers two scenarios regarding the legality of additional tax penalties, serving as a critical resource for evaluating models' understanding of complex legal problems.




