Too Long; Didn't Read

We thoroughly evaluate and analyze the probing-based Pro-Cap method for hateful meme detection. Discover its performance on different datasets, understand its impact through an ablation study and baseline comparison, and gain insights into its effectiveness and stability in detecting hateful content.