以博弈论多智能体协作框架应对模型偏差的仇恨模因分类
Tackling Model Bias via Game-theoretic Multi-agent Collaboration Framework for Hateful Meme Classification
把多个 LMM 当作博弈参与者，用个体准确率与跨模型一致性构成的混合奖励引导它们收敛到合作均衡，这一思路可直接迁移到任何需要多模态模型集成或投票的视觉-语言任务（如 VQA、多模态事实性判别），用来缓解单个模型的认知偏差。
原文依据 cognitive biases in LMMs can impede effective collaboration among models
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论仇恨模因分类旨在识别包含仇恨内容的模因，在社交媒体主导的时代日益重要。大型多模态模型（LMM）显著增强了对多模态内容的理解，推动了该领域的发展。然而，LMM 中的认知偏差会阻碍模型之间的有效协作。为解决该问题，我们提出 GECO，一个博弈论多智能体协作框架，它将多个 LMM 组织为相互作用的智能体，并采用博弈论原理引导它们趋向最优的合作均衡。GECO 集成了混合奖励机制，同时纳入个体准确率与跨模型一致性，以促进收敛到一致的合作解。此外，我们实现了高效的策略学习，并引入惩罚系数以有效优化该框架并确保训练稳定性。在五个公开基准上的大量实验表明，我们的框架取得了新的最优性能。
英文原文摘要 1060 字符
Hateful meme classification aims to identify memes containing hateful content and has become increasingly important in the era of social media dominance. Large multimodal models (LMMs) have significantly enhanced the understanding of multimodal content, advancing this field. However, cognitive biases in LMMs can impede effective collaboration among models. To address this issue, we introduce GECO, a Game-theoretic multi-agEnt Collaboration framewOrk that organizes multiple LMMs into interacting agents and employs game-theoretic principles to guide them toward an optimal cooperative equilibrium. GECO integrates a mixed bonus scheme, incorporating both individual accuracy and cross-model agreement to promote convergence toward a consistent cooperative solution. In addition, we implement efficient policy learning and introduce a penalty coefficient to optimize the framework effectively and ensure training stability. Extensive experiments on five publicly available benchmarks demonstrate that our framework achieves new state-of-the-art performance.