sample of publications conference contributions Training Implicit Generative Models via an Invariant Statistical Loss. 1-29. 2024