LongCite-45k is an open source dataset launched by Tsinghua University in 2024, which aims to improve the credibility and verifiability of large language models (LLMs) in long text question answering tasks.LongCite: Enabling LLMs to Generate Fine-grained Citations in Long-context QA". The dataset contains 44,600 high-quality question-answering data with sentence-level citations, supports long text processing with a maximum length of 128k tokens, and enables users to verify the accuracy of the model's answers by generating fine-grained sentence-level citations.

Citation

@article{zhang2024longcite, title = {LongCite: Enabling LLMs to Generate Fine-grained Citations in Long-context QA} author={Jiajie Zhang and Yushi Bai and Xin Lv and Wanjun Gu and Danqing Liu and Minhao Zou and Shulin Cao and Lei Hou and Yuxiao Dong and Ling Feng and Juanzi Li}, journal={arXiv preprint arXiv:2409.02897}, year={2024} }