AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Towards Deeper Deep Reinforcement Learning with Spectral Normalization

Johan Bjorck, Carla P Gomes, Kilian Q Weinberger

NeurIPS 2021 · Reinforcement learning › Policy gradient and actor-critic · approach: Stabilizing deep actor-critic · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Add spectral normalization to modern heads (ablation)DRQv2 (DDPG) · walker walkreturn: 6.10 $\pm$ 9.08 -> 739.94 $\pm$ 20.88Improved
Replace MLP heads with modern architecture, without spectral normalization (ablation)DRQv2 (DDPG) · walker walkreturn: 477.40 $\pm$ 117.96 -> 6.10 $\pm$ 9.08Hurt
Add spectral normalization to modern heads (ablation)DRQv2 (DDPG) · hopper standreturn: 2.95 $\pm$ 0.99 -> 796.60 $\pm$ 85.04Improved
Replace MLP heads with modern architecture, without spectral normalization (ablation)DRQv2 (DDPG) · hopper standreturn: 403.82 $\pm$ 123.85 -> 2.95 $\pm$ 0.99Hurt
Add spectral normalization to modern heads (ablation)DRQv2 (DDPG) · reacher easyreturn: 77.22 $\pm$ 19.95 -> 801.83 $\pm$ 66.31Improved
Add spectral normalization to modern heads (ablation)DRQv2 (DDPG) · cup catchreturn: 107.46 $\pm$ 21.43 -> 975.32 $\pm$ 1.3Improved
Add spectral normalization to modern heads (ablation)DRQv2 (DDPG) · cartpole sparsereturn: 11.27 $\pm$ 0.16 -> 1000.00 $\pm$ 0.00Improved
Use modern architecture without spectral normalization instead of MLP heads (ablation)DRQ (SAC) · DeepMind Control Suite tasksreturnMixed
Increase MLP hidden layers from two to four (ablation)DRQ (SAC) · DeepMind Control Suite tasksreturnMixed
Apply spectral normalization to modern networks (ablation)DRQ (SAC) · DeepMind Control Suite tasksreturnImproved
Increase modern network depth while using spectral normalization (ablation)DRQ (SAC) · cartpole swingup sparsereturnImproved
Increase architecture capacity from two-hidden-layer MLP to modern network with SNDRQ (SAC) · DeepMind Control Suite tasksreturn: slightly more than 141Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy