Datasets:

hongli-zhan
/

SPRI-SFT-dolly

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

Dataset Card for SPRI-SFT-dolly (ICML 2025)

Paper: SPRI: Aligning Large Language Models with Context-Situated Principles (Published in ICML 2025)
Authors: Hongli Zhan, Muneeza Azmat, Raya Horesh, Junyi Jessy Li, Mikhail Yurochkin
Shared by: Hongli Zhan
Arxiv Link: arxiv.org/abs/2502.03397

Citation

If you used our dataset, please cite our paper:

@inproceedings{zhan2025spri,
  title     = {SPRI: Aligning Large Language Models with Context-Situated Principles},
  author    = {Zhan, Hongli and Azmat, Muneeza and Horesh, Raya and Li, Junyi Jessy and Yurochkin, Mikhail},
  booktitle = {Proceedings of the 42nd International Conference on Machine Learning},
  year      = {2025}
}