作者位次 #4/4ICML 2026Spotlight· Social Aspects / Alignment
Position: Measuring Human Preferences in RLHF is a Social Science Problem
Bijean Ghafouri · Eun Cheol Choi · Priyanka Dey · Emilio Ferrara
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。
Bijean Ghafouri · Eun Cheol Choi · Priyanka Dey · Emilio Ferrara
Amin Banayeeanzade · Ala N. Tak · Fatemeh Bahrani · Anahita Bolourani · Leonardo Blas · Emilio Ferrara · Jonathan Gratch · Sai Praneeth Karimireddy
The ability to control LLMs’ emulated emotional states and personality traits is an essential step in enabling rich, human-centered interactions in socially interactive settings. We introduce PsySET, a Psychologically-informed benchmark to evaluate LLM Steering Effectiveness and Trustworthiness a...
Bingxin Xu · Yuzhang Shang · Binghui Wang · Emilio Ferrara
Vision-Language-Action (VLA) models are increasingly deployed in safety-critical robotic applications, yet their security vulnerabilities remain underexplored. We identify a fundamental security flaw in modern VLA systems: the combination of action chunking and delta pose representations creates...
Emily Chen · Alexander J. Bisberg · Dmitri Williams · Magy Seif El-Nasr · Emilio Ferrara