Position: LLM Unlearning Benchmarks are Weak Measures of Progress

Thaker, P.; Hu, S.; Kale, N.; Maurya, Y.; Wu, Z.S.; Smith, V.

by P. Thaker, S. Hu, N. Kale, Y. Maurya, Z.S. Wu and V. Smith

Reference:

Position: LLM Unlearning Benchmarks are Weak Measures of Progress P. Thaker, S. Hu, N. Kale, Y. Maurya, Z.S. Wu and V. SmithIn 2025 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML), 2025.

Bibtex Entry:

@inproceedings{thaker2025position,
  title={Position: {LLM} Unlearning Benchmarks are Weak Measures of Progress},
  author={P. Thaker and S. Hu and N. Kale and Y. Maurya and Z.S. Wu and V. Smith},
  booktitle={2025 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML)},
  pages={520--533},
  year={2025},
  organization={IEEE},
url=https://arxiv.org/pdf/2410.02879
}