{"url":"/dataset/mm-safetybench","name":"MM-SafetyBench","full_name":null,"description_markdown":"The **MultiModal Safety Benchmark (MM-SafetyBench)** is a comprehensive framework designed for conducting safety-critical evaluations of Multimodal Large Language Models (MLLMs). It addresses the security concerns surrounding MLLMs, which can be compromised by query-relevant images, as if the text query itself were malicious¹².\r\n\r\nHere's a brief overview of MM-SafetyBench:\r\n- **Purpose**: It aims to evaluate the vulnerability of MLLMs to adversarial attacks that use images to manipulate model responses.\r\n- **Dataset**: The benchmark includes a dataset with 13 scenarios, resulting in a total of 5,040 text-image pairs.\r\n- **Evaluation**: It has been used to assess the safety of 12 state-of-the-art MLLMs, revealing their susceptibility to image-based manipulations¹.\r\n- **Significance**: The findings from MM-SafetyBench highlight the need for improved safety measures in open-source MLLMs to protect against potential malicious exploits².\r\n\r\n(1) GitHub - isXinLiu/MM-SafetyBench. https://github.com/isXinLiu/MM-SafetyBench.\r\n(2) MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large .... https://arxiv.org/abs/2311.17600.\r\n(3) MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large .... https://paperswithcode.com/paper/query-relevant-images-jailbreak-large-multi.\r\n(4) Official github repo for SafetyBench, a comprehensive benchmark to .... https://github.com/thu-coai/SafetyBench.\r\n(5) undefined. https://doi.org/10.48550/arXiv.2311.17600.","description_withheld":null,"homepage":"https://isxinliu.github.io/Project/MM-SafetyBench","introduced_date":"2023-11-29","introduced_date_note":null,"introduced_by":{"paper":"/paper/query-relevant-images-jailbreak-large-multi","title":"MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models","first_author":"Xin Liu","url":null},"license":null,"modalities":[],"tasks":[],"languages":[],"variants":["MM-SafetyBench"],"data_loaders":[],"num_papers_in_archive":36,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[],"papers_with_a_benchmark_row":[],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":0,"samples_harvested":0,"samples_ran":0,"samples_unverified":0,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}