作者位次 #3/6ICLR 2026Poster· Social Aspects / Fairness, Equity, Justice and Safety
Safety Instincts: LLMs Learn to Trust Their Internal Compass for Self-Defense
Guobin Shen · Dongcheng Zhao · Haibo Tong · Jindong Li · Feifei Zhao · Yi Zeng
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。
Guobin Shen · Dongcheng Zhao · Haibo Tong · Jindong Li · Feifei Zhao · Yi Zeng
Haibo Tong · Zeyang Yue · Feifei Zhao · Erliang Lin · Lu Jia · Ruolin Chen · Yinqian Sun · Qian Zhang · Yi Zeng
Whether Large Language Models (LLMs) truly possess human-like Theory of Mind (ToM) capabilities has garnered increasing attention. However, existing benchmarks remain largely restricted to narrow paradigms like false belief tasks, failing to capture the full spectrum of human cognitive mechanisms...