Research map / Reinforcement learning / Offline RL
Revisiting the Minimalist Approach to Offline 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 actor penalty (BC term) (ablation) | ReBRAC · D4RL all domains | normalized score: 11.4 -> 72.2 | Helped |
| add critic penalty (ablation) | ReBRAC · D4RL all domains | normalized score: 71.5 -> 72.2 | No clear effect |
| decouple actor and critic penalty coefficients (ablation) | ReBRAC · D4RL all domains | normalized score: 71.6 -> 72.2 | Helped |
| use deeper networks (extra layers) (ablation) | ReBRAC · D4RL all domains | normalized score: 61.9 -> 72.2 | Helped |
| use deeper networks (extra layers) (ablation) | ReBRAC · Adroit | normalized score: 59.0 -> 58.0 | No clear effect |
| add larger batch to ReBRAC (ablation) | ReBRAC · AntMaze | normalized score: 75.3 -> 41.0 | Hurt |
| use large batch only on Gym-MuJoCo (vs smaller) (ablation) | ReBRAC · Gym-MuJoCo | normalized score: 75.9 -> 80.6 | Helped |
| change discount factor gamma for AntMaze (ablation) | ReBRAC · AntMaze | normalized score: 21.0 -> 75.3 | Helped |
| add large batch to TD3+BC (ablation) | TD3+BC · D4RL all domains | normalized score: 52.2 -> 7.9 | Hurt |
| add layer normalization to TD3+BC (ablation) | TD3+BC · D4RL all domains | normalized score: 52.2 -> 60.2 | Helped |
| add extra layers to TD3+BC (ablation) | TD3+BC · D4RL all domains | normalized score: 52.2 -> 61.9 | Helped |
| tune TD3+BC hyperparameters (ablation) | TD3+BC · D4RL all domains | normalized score: 52.2 -> 58.3 | Helped |
Closest papers by meaning
- A Clean Slate for Offline Reinforcement Learning (NeurIPS 2025 · similarity 0.94)
- Simple Ingredients for Offline Reinforcement Learning (ICML 2024 · similarity 0.91)
- Dispelling the Mirage of Progress in Offline MARL through Standardised Baselines and Evaluation (NeurIPS 2024 · similarity 0.90)
- Compositional Conservatism: A Transductive Approach in Offline Reinforcement Learning (ICLR 2024 · similarity 0.90)
- Fewer May Be Better: Enhancing Offline Reinforcement Learning with Reduced Dataset (ICLR 2025 · similarity 0.90)
- Reward-Consistent Dynamics Models are Strongly Generalizable for Offline Reinforcement Learning (ICLR 2024 · similarity 0.90)
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.