作者位次 #3/6ICLR 2026Poster· Deep Learning / Robustness
Tree-based Dialogue Reinforced Policy Optimization for Red-Teaming Attacks
Ruohao Guo · Afshin Oroojlooyjadid · Roshan Sridhar · Miguel Ballesteros · Alan Ritter · Dan Roth
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Ruohao Guo · Afshin Oroojlooyjadid · Roshan Sridhar · Miguel Ballesteros · Alan Ritter · Dan Roth
Xiaohan Zou · Roshan Sridhar · Mohammadtaher Safarzadeh · Dan Roth
The reliability of VLM-as-a-Judge is critical for the automatic evaluation of vision-language models (VLMs). Despite recent progress, our analysis reveals that VLM-as-a-Judge often pays limited attention to the image when making decisions. Instead, they often blindly favor the more informative an...