4 papers
Prompt-Level Reward Specifications for Open-Ended Post-Training
Zijun Weng, Xiaohui Hu, Shuangyong Song +3
Open-ended post-training benefits from rewards that make prompt-specific success conditions explicit, rather than relying only on post-hoc scalar scores. In instruction following,…
Steering the Verifiability of Multimodal AI Hallucinations
Jianhong Pang, Ruoxi Cheng, Ziyi Ye +4
AI applications driven by multimodal large language models (MLLMs) are prone to hallucinations and pose considerable risks to human users. Crucially, such hallucinations are not eq…
LLMs Deceive Unintentionally: Emergent Misalignment in Dishonesty from Misaligned Samples to Biased Human-AI Interactions
Xuhao Hu, Peng Wang, Xiaoya Lu +3
Previous research has shown that LLMs finetuned on malicious or incorrect completions within narrow domains (e.g., insecure code or incorrect medical advice) can become broadly mis…
VLSBench: Unveiling Visual Leakage in Multimodal Safety
Xuhao Hu, Dongrui Liu, Hao Li +2
Safety concerns of Multimodal large language models (MLLMs) have gradually become an important problem in various applications. Surprisingly, previous works indicate a counterintui…