Variancereduced Gradient Estimation Via Noisereuse In Online Evolution Strategies

# Variance-Reduced Gradient Estimation via Noise-Reuse in Online Evolution Strategies

**arXiv ID:** 2304.12180
**Authors:** Oscar Li, James Harrison, Jascha Sohl-Dickstein, Virginia Smith, Luke Metz
**Published:** 2023-04-21T17:53:05Z
**Abstract:**
Unrolled computation graphs are prevalent throughout machine learning but present challenges to automatic differentiation (AD) gradient estimation methods when their loss functions exhibit extreme local sensitivtiy, discontinuity, or blackbox characteristics. In such scenarios, online evolution strategies methods are a more capable alternative, while being more parallelizable than vanilla evolution strategies (ES) by interleaving partial unrolls and gradient updates. In this work, we propose a general class of unbiased online evolution strategies methods. We analytically and empirically characterize the variance of this class of gradient estimators and identify the one with the least variance, which we term Noise-Reuse Evolution Strategies (NRES). Experimentally, we show NRES results in faster convergence than existing AD and ES methods in terms of wall-clock time and number of unroll steps across a variety of applications, including learning dynamical systems, meta-training learned optimizers, and reinforcement learning.

## Skill Description

This skill is generated from the arXiv paper: Variance-Reduced Gradient Estimation via Noise-Reuse in Online Evolution Strategies (2304.12180).

## How to Use

[To be filled in by the user or by future automation]

## References

- [arXiv:2304.12180](http://arxiv.org/abs/2304.12180v2)