影响力指数

-/100

发表论文3 篇

平均评分3.9

年均产出1.5 篇/年

Canaan Yung

PhD student@University of Melbourne·澳大利亚·OpenReview

研究方向

Large Language Models

Geometry-Guided Adversarial Prompt Detection via Curvature and Local Intrinsic Dimension

ICLR 2026Rejected

Characterising Universal Jailbreak Features and Refusal Direction in LLMs

ICLR 2026Withdrawn

CURVALID: A Geometrically-guided Adversarial Prompt Detection

ICLR 2025Rejected

合作者 (3)

Christopher Leckie

博士导师3 篇

博士导师3 篇

Sarah Monazam Erfani

博士导师3 篇