BAM：用不对称二维高斯让可解释表格预测真正可审计
BAM: Bivariate Additive Models for Interpretable Tabular Prediction
EBM 这类可解释模型的形状函数经常复杂到没法看，BAM 把它们换成二维高斯，精度不差还真能拿去对着领域知识核验，做表格建模的可以看看。
论文提出 Bivariate Additive Models（BAM），属于 GA²M 的一类变体，把每个成对形状函数约束为不对称二维高斯。相比 EBM 和 NODE-GAM 学出的复杂等高线，BAM 的每个组件可画成单峰等高线图，特征对在训练中自动选择。在六个表格基准上，BAM 的预测精度与标准 GAM 相当，接近 GA²M。论文还给出了精度与需检查组件数量之间的权衡关系。
BAM: Bivariate Additive Models for Interpretable Tabular Prediction
Generalized Additive Models with pairwise interactions (GA$^2$Ms) are regarded as glass-box models, since each pairwise shape function can be visualized as a contour plot. In practice, the shape functions learned by methods such as EBM or NODE-GAM are so complex (irregular contours and sharp local fluctuations) that a practitioner cannot meaningfully interpret them or check them against domain knowledge. Auditability in this stronger sense requires substantially simpler shape functions. We argue that such simplification can match the accuracy of standard GAMs if the right feature pairs are captured. We propose Bivariate Additive Models (BAMs), a class of GA$^2$Ms in which every pairwise shape function is constrained to a simple parametric form, instantiated as an asymmetric bivariate Gaussian. Each component is visualizable as a unimodal contour map, and its feature pair is selected automatically during training. Experiments on six tabular benchmarks show that BAM attains predictive accuracy comparable to standard GAMs and approaches that of GA$^2$Ms. We also show the trade-off between accuracy and the number of components to inspect.