Beyond Translation: Cross-Cultural Meme Transcreation with Vision-Language Models

Yuming Zhao Peiyi Zhang Oana Ignat · 2026

Read Beyond Translation: Cross-Cultural Meme Transcreation with Vision-Language Models by Yuming Zhao Peiyi Zhang Oana Ignat (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-02510.txt.

Description

Yuming Zhao and 2 co-authors study how to remake a meme for another culture without killing the joke, in a 2026 arXiv paper. Memes depend on culture-specific references, so direct translation fails. Transcreation aims to preserve communicative intent and humor while swapping in references the new audience knows.

The team proposes a hybrid framework based on vision-language models. They also build a large bidirectional dataset of Chinese and United States memes. Human judgments and automated evaluation cover thousands of meme pairs in both directions. Current models manage transcreation only to a limited extent. A clear asymmetry appears: adapting United States memes for China scored consistently higher than the reverse. The authors identify which humor and design elements transfer and which resist, and they propose an evaluation framework for cross-cultural multimodal generation.

Catalog record

CREATORYuming Zhao Peiyi Zhang Oana Ignat
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE53 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-02510

Browse the full library · About the Tower of Babel