Research map / Reinforcement learning / Visual and representation RL
Quantifying Generalization in Reinforcement Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use the deeper and wider IMPALA-CNN architecture instead of Nature-CNN | PPO agents · CoinRun | test performance | Helped |
| Increase architecture size to IMPALA-Large, with 5 residual blocks and twice as many channels | IMPALA-CNN · CoinRun | test performance | Helped |
| Add L2 regularization (ablation) | IMPALA-CNN · CoinRun, 500 training levels | test performance | Helped |
| Add dropout (ablation) | IMPALA-CNN · CoinRun, 500 training levels | test performance | Helped |
| Add modified Cutout data augmentation (ablation) | IMPALA-CNN · CoinRun, 500 training levels | test performance | Helped |
| Add batch normalization after every convolutional layer (ablation) | IMPALA-CNN · CoinRun, 500 training levels | test performance | Helped |
| Increase environmental stochasticity using epsilon-greedy action selection (ablation) | PPO agent · CoinRun, 500 training levels | test performance | Helped |
| Increase policy stochasticity by raising the entropy bonus (ablation) | PPO agent · CoinRun, 500 training levels | test performance | Helped |
| Combine data augmentation, batch normalization, and L2 regularization (ablation) | IMPALA-CNN · CoinRun, 500 training levels | test performance | Helped |
| Combine epsilon-greedy action selection and high entropy bonuses with other regularization (ablation) | IMPALA-CNN · CoinRun, 500 training levels | test performance | Hurt |
| Increase the number of training levels (ablation) | IMPALA-CNN with LSTM · RandomMazes | test performance | Helped |
| Increase the number of training levels (ablation) | IMPALA-CNN with LSTM · CoinRun-Platforms | test performance | Helped |
Closest papers by meaning
- Network Randomization: A Simple Technique for Generalization in Deep Reinforcement Learning (ICLR 2020 · similarity 0.90)
- Instance-based Generalization in Reinforcement Learning (NeurIPS 2020 · similarity 0.90)
- On the Power of Pre-training for Generalization in RL: Provable Benefits and Hardness (ICML 2023 · similarity 0.90)
- The Generalization Gap in Offline Reinforcement Learning (ICLR 2024 · similarity 0.89)
- What Effects the Generalization in Visual Reinforcement Learning: Policy Consistency with Truncated Return Prediction (AAAI 2024 · similarity 0.88)
- Zero-Shot Context Generalization in Reinforcement Learning from Few Training Contexts (NeurIPS 2025 · similarity 0.88)
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.