CopyBench: Measuring Literal and Non-Literal Reproduction of Copyright-Protected Text in Language Model Generation

Tong Chen1 , Akari Asai1* , Niloofar Mireshghallah1* , Sewon Min1 , James Grimmelmann2,3 ,
Yejin Choi1,4 , Hannaneh Hajishirzi1,4 , Luke Zettlemoyer1 , Pang Wei Koh1,4
1University of Washington, 2Cornell University, 3Cornell Law School, 4Allen Institute for AI
Paper Code Data Leaderboard

Dataset Viewer

BibTeX

@misc{chen2024copybenchmeasuringliteralnonliteral,
          title={CopyBench: Measuring Literal and Non-Literal Reproduction of Copyright-Protected Text in Language Model Generation}, 
          author={Tong Chen and Akari Asai and Niloofar Mireshghallah and Sewon Min and James Grimmelmann and Yejin Choi and Hannaneh Hajishirzi and Luke Zettlemoyer and Pang Wei Koh},
          year={2024},
          eprint={2407.07087},
          archivePrefix={arXiv},
          primaryClass={cs.CL},
          url={https://arxiv.org/abs/2407.07087}, 
    }