标签
This paper introduces CLAM, a method for estimating localized causal effects from coarse-resolution data by jointly learning causal mechanisms and a disaggregation mapping, with applications in public health and environmental policy.