mlsanityData to 7 Oct 2026

Guiding Evolutionary Strategies by Differentiable Robot Simulators

arXiv 2110.00438 ↗First posted 1 Oct 2021cs.RO cs.AI cs.LGcited by 1

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 11 of its 17 references that are in the map. First posted on arXiv 1 Oct 2021.

Builds on

11 of 17 references.

References · 11

  1. 2020Learning to Simulate Complex Physics with Graph Networkscited by 867
  2. 2017Evolution Strategies as a Scalable Alternative to Reinforcement Learningcited by 555
  3. 2019DiffTaichi: Differentiable Programming for Physical Simulationcited by 248
  4. 2016A Differentiable Physics Engine for Deep Learning in Roboticscited by 107
  5. 2017Improving Exploration in Evolution Strategies for Deep Reinforcement Learning via a Population of Novelty-Seeking Agentscited by 107
  6. 2018Evolution-Guided Policy Gradient in Reinforcement Learningcited by 78
  7. 2018Simple random search provides a competitive approach to reinforcement learningcited by 75
  8. 2020Scalable Differentiable Physics for Learning and Controlcited by 72
  9. 2018CEM-RL: Combining evolutionary and gradient-based methods for policy searchcited by 53
  10. 2020On the distance between two neural networks and the stability of learningcited by 46
  11. 2018Back to Basics: Benchmarking Canonical Evolution Strategies for Playing Ataricited by 26

Built on by

Cited by year

Citing papers per year: 2023: 1.

1 paper.

  1. 2023High-performance real-world optical computing trained by in situ gradient-based model-free optimizationcited by 1