You Never Know: Quantization Induces Inconsistent Biases in Vision-Language Foundation Models

Open in new window