LIMITGEN
收藏资源简介:
LIMITGEN是一个全面的数据集,旨在评估大型语言模型(LLM)在识别科学研究中论文局限性的能力。该数据集由两个子集组成:LIMITGEN-Syn,一个通过控制扰动高质量论文创建的合成数据集;LIMITGEN-Human,一个收集真实人类撰写的局限性的数据集。LIMITGEN的创建是为了帮助LLM系统在研究论文中生成局限性,以便它们能够提供更具体和建设性的反馈。该数据集的创建过程涉及对科学论文的全面分类,以及通过控制扰动和人类审查来创建合成数据集和收集真实数据。LIMITGEN的应用领域是科学研究,旨在帮助研究人员识别和解决研究中的局限性,以促进科学进步。
LIMITGEN is a comprehensive dataset designed to evaluate the ability of Large Language Models (LLMs) to identify the limitations of scientific research papers. This dataset consists of two subsets: LIMITGEN-Syn, a synthetic dataset created via controlled perturbation of high-quality scientific papers; and LIMITGEN-Human, a dataset of authentic limitations written by human researchers. The development of LIMITGEN aims to assist LLM systems in generating limitation sections for research papers, enabling them to provide more specific and constructive feedback. The construction process of LIMITGEN involves comprehensive categorization of scientific papers, as well as the creation of the synthetic dataset through controlled perturbation and the curation of real human-authored limitation content via human review. The application scope of LIMITGEN is scientific research, with the goal of helping researchers identify and address limitations in their studies to advance scientific progress.

- 1Can LLMs Identify Critical Limitations within Scientific Research? A Systematic Evaluation on AI Research Papers耶鲁大学 · 2025年



