影响力指数

43.68/100

前 11.8%

全站排名 #7,613

发表论文3 篇

平均评分6.1

年均产出1.5 篇/年

Kristina Nikolić

PhD student@ETHZ - ETH Zurich·瑞士·OpenReview

研究方向

AI Safety · Adversarial Attacks · Machine Learning

Strategic Dishonesty Can Undermine AI Safety Evaluations of Frontier LLMs

ICLR 2026Poster

Modal Aphasia: Can Unified Multimodal Models Describe Images From Memory?

ICLR 2026Poster

The Jailbreak Tax: How Useful are Your Jailbreak Outputs?

ICML 2025Spotlight

合作者 (13)

Florian Tramèr

博士导师2 篇

Alexander Panfilov

Evgenii Kortukov