🧪 Test?View on arXiv
IRON-Meme: A Benchmark for Structured Understanding of Identities and Roles in Online Meme Narratives
Author1, Author2, Author3, Author4, Author5
multimodalreasoningdatasetsocial identities
2610.33097
Builder Relevance
1h ago70%
Abstract
The paper introduces IRON-Meme, a benchmark for analyzing political memes through a structured framework that connects social identities, viewpoints, and narrative roles.
Reality Card
Core Claim
The introduction of IRON-Meme provides a comprehensive human-annotated dataset that reveals the interconnectedness of identities, roles, and templates in political memes.
Method / Result
Evaluation of nine vision-language models shows that frontier-scale and reasoning models achieve the highest performance.
Limitations
Accuracy on social identities and narrative roles decreases for less frequent labels, which may affect generalizability.
Paper to code
Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.
No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.