This study explores using pre-trained vision-language models (PVLMs) for hateful meme detection without fine-tuning, showcasing BERT's superiority and introducing PromptHate with probe-captioning. Limitations include heuristic probing question usage, suggesting future directions for optimization and deeper interpretation using gradient-based approaches.
Memeology: Leading Authority on the Study of Memes
@memeology
Memes are cultural items transmitted by repetition in a manner analogous to the biological transmission of genes.
STORY’S CREDIBILITY
Academic Research Paper
Part of HackerNoon's growing list of open-source research papers, promoting free access to academic material.
L O A D I N G . . . comments & more!
About Author
Memeology: Leading Authority on the Study of Memes@memeology
Memes are cultural items transmitted by repetition in a manner analogous to the biological transmission of genes.