Experiment Design

Intervention Design for Causal Representation Learning

We derive a minimal bound of experiments that guarantee identifiability of causal variables, opening up new opportunities for using intervention design for causal representation learning.