用偏好优化缓解概念标签错误对模型性能的损害
Addressing Concept Mislabeling in Concept Bottleneck Models Through Preference Optimization
- 提出基于直接偏好优化的新损失函数CPO，降低标签噪声影响
- 在三个真实数据集上，CPO比BCE提升性能，噪声下表现更稳
- 适合关注可解释AI、标签质量不佳场景的研究者
概念瓶颈模型（CBMs）通过约束决策基于人类可理解的概念来提升AI系统的可信度。然而，这类模型通常假设数据集包含准确的概念标签，这一假设在实践中常被违反，我们发现这会导致性能显著下降（某些情况下达25%）。为此，我们引入概念偏好优化（CPO）目标，一种基于直接偏好优化的新损失函数，能有效缓解概念标签错误对CBM性能的负面影响。我们分析了CPO的关键性质，表明其直接优化概念后验分布，并与二元交叉熵（BCE）对比，证明CPO对概念噪声具有内在鲁棒性。我们在三个真实世界数据集上实证验证了该分析，结果表明，无论是否有额外标签噪声，CPO始终优于BCE。代码已开源。
原文摘要 · Abstract (English)
Concept Bottleneck Models (CBMs) propose to enhance the trustworthiness of AI systems by constraining their decisions on a set of human-understandable concepts. However, CBMs typically assume that datasets contain accurate concept labels-an assumption often violated in practice, which we show can significantly degrade performance (by 25% in some cases). To address this, we introduce the Concept Preference Optimization (CPO) objective, a new loss function based on Direct Preference Optimization, which effectively mitigates the negative impact of concept mislabeling on CBM performance. We provide an analysis of key properties of the CPO objective, showing it directly optimizes for the concept's posterior distribution, and contrast it against Binary Cross Entropy (BCE), demonstrating that CPO is inherently less sensitive to concept noise. We empirically confirm our analysis by finding that CPO consistently outperforms BCE on three real-world datasets, both with and without added label noise. We make our code available on Github.
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。