Research map / Reinforcement learning / MDPs and planning
Instance-based 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 |
|---|---|---|---|
| Add ℓ² regularization | base · CoinRun test | episode return: 7.47±.25 -> 7.49±.22 | No clear effect |
| Add ℓ² regularization and Cutout | base · CoinRun test | episode return: 7.47±.25 -> 7.99±.20 | Helped |
| Use IAPE instead of base training | CoinRun agent · CoinRun test | episode return: 7.47±.25 -> 8.23±.18 | Helped |
| Use IAPE instead of base training | CoinRun agent · CoinRun test | policy KL divergence to ∞-levels: 0.57±0.39 -> 0.1±0.09 | Helped |
| Add ℓ² regularization | base · CoinRun test | policy KL divergence to ∞-levels: 0.57±0.39 -> 0.19±0.15 | Helped |
| Add ℓ² regularization and Cutout | base · CoinRun test | policy KL divergence to ∞-levels: 0.57±0.39 -> 0.15±0.11 | Helped |
| Collect experience using the consensus policy rather than each subset's own policy (ablation) | IAPE · CoinRun test | episode return | Helped |
| Collect experience using the consensus policy rather than each subset's own policy (ablation) | IAPE · CoinRun test | policy agreement with the consensus policy | Helped |
Closest papers by meaning
- Why Generalization in RL is Difficult: Epistemic POMDPs and Implicit Partial Observability (NeurIPS 2021 · similarity 0.91)
- Quantifying Generalization in Reinforcement Learning (ICML 2019 · similarity 0.90)
- How Ensembles of Distilled Policies Improve Generalisation in Reinforcement Learning (NeurIPS 2025 · similarity 0.88)
- Mitigating Partial Observability in Sequential Decision Processes via the Lambda Discrepancy (NeurIPS 2024 · similarity 0.88)
- Foundation Policies with Hilbert Representations (ICML 2024 · similarity 0.88)
- Bridging State and History Representations: Understanding Self-Predictive RL (ICLR 2024 · 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.