[2604.00547] Does Unification Come at a Cost? Uni-SafeBench: A Safety Benchmark for Unified Multimodal Large Models

Computer Science > Artificial Intelligence arXiv:2604.00547 (cs) [Submitted on 1 Apr 2026] Title:Does Unification Come at a Cost? Uni-SafeBench: A Safety Benchmark for Unified Multimodal Large Models Authors:Zixiang Peng, Yongxiu Xu, Qinyi Zhang, Jiexun Shen, Yifan Zhang, Hongbo Xu, Yubin Wang, Gaopeng Gou View a PDF of the paper titled Does Unification Come at a Cost? Uni-SafeBench: A Safety Benchmark for Unified Multimodal Large Models, by Zixiang Peng and 7 other authors View PDF HTML (experimental) Abstract:Unified Multimodal Large Models (UMLMs) integrate understanding and generation capabilities within a single architecture. While this architectural unification, driven by the deep fusion of multimodal features, enhances model performance, it also introduces important yet underexplored safety challenges. Existing safety benchmarks predominantly focus on isolated understanding or generation tasks, failing to evaluate the holistic safety of UMLMs when handling diverse tasks under a unified framework. To address this, we introduce Uni-SafeBench, a comprehensive benchmark featuring a taxonomy of six major safety categories across seven task types. To ensure rigorous assessment, we develop Uni-Judger, a framework that effectively decouples contextual safety from intrinsic safety. Based on comprehensive evaluations across Uni-SafeBench, we uncover that while the unification process enhances model capabilities, it significantly degrades the inherent safety of the underlying ...