SOURCE-LINKED INTELLIGENCE

MemeCULT-1K: Benchmarking South Asian Cultural Context and Humor Understanding of Multimodal Models

arXiv · AI, language, vision and robotics · article · Sep 1, 2026 · UTC

Meme understanding goes beyond recognizing visual content or literal text; it requires implicit cultural knowledge and pragmatic inference that most vision-language models still lack. We introduce MemeCULT-1K, a multilingual benchmark of 1,000 South Asian memes in Bengali, English, and Hindi, where each meme is paired with a cultural context note and three human-written explanations, along with a supplementary set of 54 Bengali regional dialect memes. We evaluate thirteen popular Vision Language Models (VLMs) under two settings: meme-only and context-aware. Providing minimal cultural context y

Read original source ↗ Open in workspace

recordType

paper

region

Global