AdamMeme is an adaptive, agent-based evaluation framework that iteratively refines meme text to expose model-specific weaknesses in multimodal models' understanding of meme harmfulness.
This particular meme can contribute to thenormalization of racially charged humor, making it more acceptable to dehumanize individuals based on race in a broader societal context
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
AdamMeme: Adaptively Probe the Reasoning Capacity of Multimodal Large Language Models on Harmfulness
AdamMeme is an adaptive, agent-based evaluation framework that iteratively refines meme text to expose model-specific weaknesses in multimodal models' understanding of meme harmfulness.