一作ACL 2026Poster· Natural Language Processing
The Side Effects of Being Smart: Safety Risks in MLLM s’ Multi-Image Reasoning
Renmiao Chen · Yida Lu · Shiyao Cui · Xuan Ouyang · Victor Shea-Jay Huang · Shumin Zhang · Chengwei Pan · Han Qiu · Minlie Huang
摘要
As Multimodal Large Language Models (MLLMs) acquire stronger reasoning capabilities to handle complex, multi-image instructions, this advancement may pose new safety risks. We study this problem by introducing MIR-SafetyBench, the first benchmark focused on multi-image reasoning safety, which con...