作者位次 #7/8ACL 2026Findings· Findings of ACL
S afe S teer: A Decoding-level Defense Mechanism for Multimodal Large Language Models
Xinyi Zeng · Xue Yang · Jingyuan Zhang · Huanqian Yan · Xiang Chen · Kaiwen Wei · Hankun Kang · Yu Tian
摘要
Multimodal large language models (MLLMs) are gaining increasing attention. Due to the heterogeneity of their input features, they face significant challenges in terms of jailbreak defenses. Current defense methods rely on costly fine-tuning or inefficient post-hoc interventions, limiting their ab...