Review for NeurIPS paper: Differentiable Expected Hypervolume Improvement for Parallel Multi-Objective Bayesian Optimization