Loaded DiCE: Trading off bias and variance in any-order score function gradient estimators for reinforcement learning
| Main Authors: | , , |
|---|---|
| Format: | Conference item |
| Published: |
Neural Information Processing Systems Foundation
2019
|