Shaopeng Fu
1 indexed paper
Recent (6 mo)
1With code
0Influential cites
0Benchmarked
0Publications per year
126
Top categories
ML×1Crypto×1Stats ML×1
Frequent co-authors
Di Wang1×
Research Timeline
2026
Understanding and Improving Continuous Adversarial Training for LLMs via In-context Learning Theory
This paper theoretically analyzes Continuous Adversarial Training (CAT) for LLMs using In-context Learning (ICL) theory, proving that embedding space perturbations effectively enhance robustness against token-space jailbreaks and proposing a singular value regularization method for improvement.
Highlighted terms show continued research focus across papers