Tackling Model Bias via Game-theoretic Multi-agent Collaboration Framework for Hateful Meme Classification

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 tackle 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 further integrates a mixed bonus scheme with both individual accuracy and cross-model agreement, which together drive the system 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 public datasets demonstrate that our framework achieves new state-of-the-art performance. We release our code. †