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
8.82/100
超过 27.6%
全站排名 #46,627
发表论文6 篇
平均评分
年均产出2.0 篇/年
Kellin Pelrine
研究方向
Generative agents · LLMs · simulation · digital twin · copilot · tutor · web search · AI Safety · persuasion · manipulation · jailbreaks · robustness
17
It's the Thought that Counts: Evaluating the Attempts of Frontier LLMs to Persuade on Harmful Topics
ICLR 2026Rejected
通讯16
TamperBench: Systematically Stress-Testing LLM Safety Under Fine-Tuning and Tampering
ICLR 2026Rejected
7
Accidental Vulnerability: Factors in Fine-Tuning that Shift Model Safeguards
ICLR 2026Rejected
三作