# ExPO-HM: Learning to Explain-then-Detect for Hateful Meme Detection

arXiv: https://arxiv.org/abs/2510.08630 · HTML: https://mlsanity.com/paper/2510.08630/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 29 of its 41 references that are in the map.
- First posted on arXiv 8 Oct 2025.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 29 of 29)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [Flamingo: a Visual Language Model for Few-Shot Learning](https://mlsanity.com/paper/2204.14198/) (2022)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [The Hateful Memes Challenge: Detecting Hate Speech in Multimodal Memes](https://mlsanity.com/paper/2005.04790/) (2020)
- [The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models](https://mlsanity.com/paper/2505.22617/) (2025)
- [HelpSteer: Multi-attribute Helpfulness Dataset for SteerLM](https://mlsanity.com/paper/2311.09528/) (2023)
- [MOMENTA: A Multimodal Framework for Detecting Harmful Memes and Their Targets](https://mlsanity.com/paper/2109.05184/) (2021)
- [Visual Program Distillation: Distilling Tools and Programmatic Reasoning into Vision-Language Models](https://mlsanity.com/paper/2312.03052/) (2023)
- [Hate-CLIPper: Multimodal Hateful Meme Classification based on Cross-modal Interaction of CLIP Features](https://mlsanity.com/paper/2210.05916/) (2022)
- [Pro-Cap: Leveraging a Frozen Vision-Language Model for Hateful Meme Detection](https://mlsanity.com/paper/2308.08088/) (2023)
- [Detecting Propaganda Techniques in Memes](https://mlsanity.com/paper/2109.08013/) (2021)
- [Decoding the Underlying Meaning of Multimodal Hateful Memes](https://mlsanity.com/paper/2305.17678/) (2023)
- [HARE: Explainable Hate Speech Detection with Step-by-Step Reasoning](https://mlsanity.com/paper/2311.00321/) (2023)
- [MemeCLIP: Leveraging CLIP Representations for Multimodal Meme Classification](https://mlsanity.com/paper/2409.14703/) (2024)
- [Towards Low-Resource Harmful Meme Detection with LMM Agents](https://mlsanity.com/paper/2411.05383/) (2024)
- [Improving Hateful Meme Detection through Retrieval-Guided Contrastive Learning](https://mlsanity.com/paper/2311.08110/) (2023)
- [Mapping Memes to Words for Multimodal Hateful Meme Classification](https://mlsanity.com/paper/2310.08368/) (2023)
- [Towards Explainable Harmful Meme Detection through Multimodal Debate between Large Language Models](https://mlsanity.com/paper/2401.13298/) (2024)
- [Modularized Networks for Few-shot Hateful Meme Detection](https://mlsanity.com/paper/2402.11845/) (2024)
- [Demystifying Hateful Content: Leveraging Large Multimodal Models for Hateful Meme Detection with Explainable Decisions](https://mlsanity.com/paper/2502.11073/) (2025)
- [TotalDefMeme: A Multi-Attribute Meme dataset on Total Defence in Singapore](https://mlsanity.com/paper/2305.17911/) (2023)
- [Robust Adaptation of Large Multimodal Models for Retrieval Augmented Hateful Meme Detection](https://mlsanity.com/paper/2502.13061/) (2025)
- [MemeIntel: Explainable Detection of Propagandistic and Hateful Memes](https://mlsanity.com/paper/2502.16612/) (2025)
- [Explainability and Hate Speech: Structured Explanations Make Social Media Moderators Faster](https://mlsanity.com/paper/2406.04106/) (2024)
- [HateSieve: A Contrastive Learning Framework for Detecting and Segmenting Hateful Content in Multimodal Memes](https://mlsanity.com/paper/2408.05794/) (2024)

## Built on by (most cited first, 8 of 8)

- [MemeLens: Multilingual Multitask VLMs for Memes](https://mlsanity.com/paper/2601.12539/) (2026)
- [Can Thinking Models Think to Detect Hateful Memes?](https://mlsanity.com/paper/2603.01225/) (2026)
- [Adapting Reinforcement Learning with Chain-of-Thought Supervision for Explainable Detection of Hateful and Propagandistic Memes](https://mlsanity.com/paper/2606.15307/) (2026)
- [I Know What You Meme, Even If it Emerged Today: Understanding Evolving Memes through Open-World Knowledge Acquisition](https://mlsanity.com/paper/2606.05316/) (2026)
- [HarmTrace: Anchor-Calibrated Decoupled Optimization for Fine-Grained Target Identification in Harmful Memes](https://mlsanity.com/paper/2608.16622/) (2026)
- [Learn Before You Judge: Progressive Knowledge-to-Decision Alignment for Explainable Hateful Meme Detection](https://mlsanity.com/paper/2609.19778/) (2026)
- [Controllable Multi-label Video Safety Detection via Adaptive Tversky Policy Optimization](https://mlsanity.com/paper/2610.02019/) (2026)
- [MemeLogos: Knowledge-Driven Retrieval-Augmented Reasoning for Harmful Meme Detection](https://mlsanity.com/paper/2610.15463/) (2026)

## Abstract (arXiv)

Hateful memes have emerged as a particularly challenging form of online abuse, motivating the development of automated detection systems. Most prior approaches rely on direct detection, producing only binary predictions. Such models fail to provide the context and explanations that real-world moderation requires. Recent Explain-then-Detect approaches, using Chain-of-Thought prompting or LMM agents, perform worse than simple SFT baselines, and even advanced post-training methods such as GRPO fail to close the gap. Our analysis identifies two key issues of such systems: important policy-relevant cues such as targets and attack types are not hypothesized by the model as a likely explanation; and the binary reward signal is insufficient to guide reasoning. To address these challenges, we propose ExPO-HM (Explain-then-Detect Policy Optimization for Hateful Memes), inspired by the training and evaluation process of human annotators. ExPO-HM combines SFT warmup, GRPO with curriculum learning, and Conditional Decision Entropy (CDE) as both metric and reward for reasoning quality. Across three hateful meme benchmarks, ExPO-HM achieves state-of-the-art performance on binary detection, fine-grained classification, and reasoning quality, with up to 15% and 17% F1 improvement over the GRPO and DPO baselines, respectively. By moving hateful meme detection from simple binary alarms to explanation-driven detection, ExPO-HM provides accurate, interpretable, and actionable moderation support. Code available at https://github.com/JingbiaoMei/ExPO-HM