Awesome
How Language Model Hallucinations Can Snowball
Muru Zhang, Ofir Press, William Merrill, Alisa Liu, Noah A. Smith
Paper: https://arxiv.org/abs/2305.13534
NOTICE: If you downloaded this dataset before the 27th of May 2023 please re-download it, as the previously-uploaded flights dataset had an issue.
The answers for each dataset are either always 'yes' or 'no':
- 'no' for the flights dataset (there is never a sequence of connecting flights)
- 'yes' for the prime dataset (all the numbers are prime)
- 'no' for the senator dataset (no senator satisfies both requirements- being from a specific state and having gone to a specific college)
If you use our datasets in your work please cite:
@misc{zhang2023language,
title={How Language Model Hallucinations Can Snowball},
author={Muru Zhang and Ofir Press and William Merrill and Alisa Liu and Noah A. Smith},
year={2023},
eprint={2305.13534},
archivePrefix={arXiv},
primaryClass={cs.CL}
}