Search papers, labs, and topics across Lattice.
1
0
3
Retrieval-augmented models beat large vision-language models like LLaVA on Bengali hateful meme detection, suggesting that fine-tuning and retrieval are crucial for low-resource languages.
