@article{safesailtowardsafinegrainedsafetylandsca, title = {Safe-SAIL: Towards a Fine-grained Safety Landscape of Large Language Models via Sparse Autoencoder Interpretation Framework}, author = {Jiaqi Weng and Han Zheng and Hanyu Zhang and Ej Zhou and Qinqin He and Jialing Tao and Hui Xue and Zhixuan Chu and Xiting Wang}, year = {2025}, eprint = {2509.18127}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2509.18127}, }