Guiding Evolutionary Strategies By Differentiable Robot Simulators
ASecurity
**arXiv ID:** 2110.00438 **Authors:** Vladislav Kurenkov, Bulat Maksudov **Published:** 2021-10-01T14:20:00Z **Abstract:** In recent years, Evolutionary Strategies were actively explored in robotic tasks for policy search as they provide a simpler alternative to reinforcement learning algorithms. However, this class of algorithms is often claimed to be extremely sample-inefficient. On the other hand, there is a growing interest in Differentiable Robot Simulators (DRS) as they potentially can ...
Installs into .claude/skills of the current project.
Sponsors
SSkills DirectorySkills Directory
Know which skills are safe — weekly.
SSkills DirectorySkills Directory
Is that skill safe to install?
Are you the author of Guiding Evolutionary Strategies By Differentiable Robot Simulators?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/hiyenwong-guiding-evolutionary-strategies-by-differentiable)
# Guiding Evolutionary Strategies by Differentiable Robot Simulators
**arXiv ID:** 2110.00438
**Authors:** Vladislav Kurenkov, Bulat Maksudov
**Published:** 2021-10-01T14:20:00Z
**Abstract:**
In recent years, Evolutionary Strategies were actively explored in robotic tasks for policy search as they provide a simpler alternative to reinforcement learning algorithms. However, this class of algorithms is often claimed to be extremely sample-inefficient. On the other hand, there is a growing interest in Differentiable Robot Simulators (DRS) as they potentially can find successful policies with only a handful of trajectories. But the resulting gradient is not always useful for the first-order optimization. In this work, we demonstrate how DRS gradient can be used in conjunction with Evolutionary Strategies. Preliminary results suggest that this combination can reduce sample complexity of Evolutionary Strategies by 3x-5x times in both simulation and the real world.
## Skill Description
This skill is generated from the arXiv paper: Guiding Evolutionary Strategies by Differentiable Robot Simulators (2110.00438).
## How to Use
[To be filled in by the user or by future automation]
## References
- [arXiv:2110.00438](http://arxiv.org/abs/2110.00438v3)