影响力指数

-/100

发表论文3 篇

平均评分5.1

年均产出1.5 篇/年

Yihan Li

PhD student@National University of Defense Technology·中国·OpenReview

RLCracker: Exposing the Vulnerability of LLM Watermarks with Adaptive RL Attacks

ICLR 2026Desk Rejected

Archilles' Heel in Semi-open LLMs: Hiding Bottom against Recovery Attacks

ICLR 2025Withdrawn

Toward Trustworthy: A Method for Detecting Fine-Tuning Origins in LLMs

ICLR 2025Rejected

合作者 (13)