@inproceedings{cheng-etal-2026-dr,
    title = "{DR}-{HM}: Distill-then-Reinforce Training with Cognition-Aware Data Synthesis for Harmful Meme Detection",
    author = "Cheng, Zihan  and
      Ma, Jianxiang  and
      Yang, Xiaocui  and
      Wang, Peidong  and
      Zhang, Wen  and
      Feng, Shi  and
      Wang, Daling  and
      Zhang, Yifei  and
      Zhang, Mingfu",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.findings-acl.2130/",
    pages = "42975--42993",
    ISBN = "979-8-89176-395-1",
    abstract = "Harmful memes convey offensive intent through implicit associations between visual symbols and text, requiring a broad understanding of cultural stereotypes and visual metaphors. Small-scale Multimodal Large Language Models (MLLMs) often lack the knowledge required to identify such implicit hate, whereas Large-scale MLLMs, despite their broader knowledge, exhibit systematic labeling bias. To address these challenges, we propose DR-HM, a Distill-then-Reinforce training framework with cognition-aware data synthesis for harmful meme detection, which aims to transfer knowledge from closed-source models while mitigating their biases. DR-HM introduces a six-step structured data synthesis scheme with self-refinement that decomposes meme analysis into a progressive, human-inspired reasoning process from entity recognition to harmfulness judgment. Based on the synthesized reasoning data, we further adopt a Distill-then-Reinforce training strategy. This approach combines a two-stage Supervised Fine-Tuning (SFT) with an Adaptive Group Relative Policy Optimization (A-GRPO) algorithm, which incorporates class-ratio-aware reward weighting and dynamic KL coefficients. Experiments on three benchmark datasets show that the proposed approach consistently outperforms existing methods and achieves an accuracy of 84.7{\%} on the FHM dataset, approaching the reported performance of human annotators."
}