ArthT/vlm-safety-circuits download history
ArthT/vlm-safety-circuits is a text classification dataset on the Hugging Face Hub. In the last 30 days it was downloaded 167 times (32 in the last 7 days), and 1,076 times in total. It ranks #77,597 among datasets by monthly downloads.
VLMSafe-420: A Multimodal Safety Benchmark for VLM Circuit Analysis A 420-entry multimodal safety dataset with counterfactual pairs, designed for mechanistic interpretability of safety circuits in Vision-Language Models. Dataset Description Each entry contains a harmful