ExPO-HM：为仇恨表情包检测学习解释-然后检测
ExPO-HM: Learning to Explain-then-Detect for Hateful Meme Detection
AI总结：
ExPO-HM通过结合SFT预热、GRPO与课程学习及CDE，提升仇恨表情包检测的解释性和准确性。
摘要：
仇恨表情包已经 emerged 为一种特别具有挑战性的在线滥用形式，这推动了自动化检测系统的开发。大多数先前的方法依赖于直接检测，仅产生二元预测。此类模型无法提供现实世界审核所需的情境和解释。最近的解释-然后检测方法，使用思维链提示或LMM代理，在简单SFT基线之上表现更差，甚至高级的后训练方法如GRPO也无法缩小差距。我们的分析确定了此类系统的两个关键问题：重要的政策相关线索如目标和攻击类型未被模型假设为可能的解释；二元奖励信号不足以指导推理。为了解决这些挑战，我们提出了ExPO-HM（为仇恨表情包的解释-然后检测策略优化），受人类标注员的训练和评估过程启发。ExPO-HM结合了SFT预热、GRPO与课程学习以及条件决策熵（CDE）作为推理质量的指标和奖励。在三个仇恨表情包基准测试中，ExPO-HM在二元检测、细粒度分类和推理质量上均取得最佳性能，相比GRPO和DPO基线，F1提升高达15%和17%。通过将仇恨表情包检测从简单的二元警报转变为基于解释的检测，ExPO-HM提供了准确、可解释且可操作的审核支持。代码可在https://github.com/JingbiaoMei/ExPO-HM上获得。
英文摘要：
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
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
