Learning Invariant Graph Representations for Out-of-Distribution Generalization (Appendix)

Neural Information Processing Systems 

In this section, we provide the proofs of Theorem 3.2 and 4.1.