影响力指数
论文质量、代表作、近期表现、广度与样本量置信度综合计算
-/100
发表论文3 篇
平均评分
年均产出3.0 篇/年
Kailong Wang
研究方向
large language model security · AI safety · AI trustworthiness
12
Evading Safety Alignment in Open-source LLM Deployment by Embedding Semantic Shift
ICLR 2026Rejected
通讯15
Circumventing Safety Alignment in Large Language Models via Embedding Space Toxicity Attenuation
ICLR 2026Rejected
通讯16
Seeing It Before It Happens: In-Generation NSFW Detection for Diffusion-Based Text-to-Image Models
ICLR 2026Withdrawn
通讯