Paper

Average versus expected causal effects under interference: Modeling assumptions and estimation bottlenecks · ByteTrail