English

Backpropagation through the Void: Optimizing control variates for black-box gradient estimation

Machine Learning 2018-02-27 v3

Abstract

Gradient-based optimization is the foundation of deep learning and reinforcement learning. Even when the mechanism being optimized is unknown or not differentiable, optimization using high-variance or biased gradient estimates is still often the best strategy. We introduce a general framework for learning low-variance, unbiased gradient estimators for black-box functions of random variables. Our method uses gradients of a neural network trained jointly with model parameters or policies, and is applicable in both discrete and continuous settings. We demonstrate this framework for training discrete latent-variable models. We also give an unbiased, action-conditional extension of the advantage actor-critic reinforcement learning algorithm.

Keywords

Cite

@article{arxiv.1711.00123,
  title  = {Backpropagation through the Void: Optimizing control variates for black-box gradient estimation},
  author = {Will Grathwohl and Dami Choi and Yuhuai Wu and Geoffrey Roeder and David Duvenaud},
  journal= {arXiv preprint arXiv:1711.00123},
  year   = {2018}
}

Comments

Published at ICLR 2018

R2 v1 2026-06-22T22:32:18.217Z