Review for NeurIPS paper: Identifying Learning Rules From Neural Network Observables