TY  - RPRT
TI  - Distillability of LLM Security Logic: Predicting Attack Success Rate of Outline Filling Attack via Ranking Regression
AU  - Tianyu Zhang
AU  - Zihang Xi
AU  - Jingyu Hua
AU  - Sheng Zhong
PY  - 2025
UR  - https://arxiv.org/abs/2511.22044
ID  - 2511.22044
ER  -