通过用户反馈引导，让图神经网络摆脱依赖表面线索的错误模式。
Overcoming Shortcut Learning in Graph Neural Networks through Active Explanation Guidance
- 用模型解释检测出误导性关联线索，再由专家提供修正反馈。
- 主动筛选高风险解释，降低人工标注和认知成本。
- 适用于各类图神经网络，提升模型在新场景下的可靠性。
图神经网络（GNN）可能无意中利用与预测无关但相关的表面线索——如某些边、节点或特征——导致其在分布外任务中可靠性下降。本文提出XIGL，一种无需修改架构的人机协同策略，用于消除此类短路行为。核心思路是：一方面，可通过分析GNN的解释发现对短路线索的依赖；另一方面，经验丰富的用户在意识到这些线索后，可提供针对性的纠正反馈，帮助模型去混淆。XIGL支持任意查询策略，为降低反馈成本，我们设计了主动学习机制，优先选择更可能暴露短路行为的解释。我们在多个GNN架构上验证了XIGL的有效性，包括现有及新提出的基于解释的策略。代码已公开。
原文摘要 · Abstract (English)
Graph Neural Networks (GNNs) can solve prediction tasks by unintentionally exploiting shortcuts---that is, edges, nodes, and features that correlate with but are not causal for the prediction---which compromise their reliability in out-of-distribution tasks. We introduce XIGL, an architecture-agnostic human-in-the-loop strategy for removing such shortcuts from GNNs. Our key insight is twofold. On the one hand, reliance on shortcuts can be detected by inspecting GNN explanations. On the other hand, once made aware of such shortcuts, sufficiently expert users can provide tailored corrective feedback, which helps deconfound the model. XIGL supports any query strategy; however, since corrective feedback can be expensive to acquire, we develop an active learning strategy for prioritizing explanations that are more likely to display shortcut behavior, lowering annotation and cognitive costs. We showcase the effectiveness of XIGL, including both existing and proposed explanation-based strategies, on several GNN architectures. Our implementation is available online.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。