Research map / Reinforcement learning / Offline RL
Efficient Offline Reinforcement Learning via Peer-Influenced Constraint
Yujia Zhang, Lin Li, Wei Wei, Jianguo Wu, Yi Ma, Jiye Liang
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add critic ensemble to PIC-TD3 (EPIC) (ablation) | PIC-TD3 · Gym-MuJoCo average | normalized score: 85.1 -> 87.8 | Improved |
| add critic ensemble to PIC-TD3 (EPIC) (ablation) | PIC-TD3 · AntMaze average | normalized score: 75.6 -> 82.9 | Improved |
| add critic ensemble to PIC-TD3 (EPIC) (ablation) | PIC-TD3 · Adroit average | normalized score: 53.8 -> 62.5 | Improved |
| replace behavior-cloning constraint with peer-influenced constraint | TD3+BC · Gym-MuJoCo average | normalized score: 70.4 -> 85.1 | Improved |
| replace behavior-cloning constraint with peer-influenced constraint | TD3+BC · Adroit average | normalized score: 49.9 -> 53.8 | Improved |
| replace behavior-cloning constraint with peer-influenced constraint | TD3+BC · AntMaze average | normalized score: 28.8 -> 75.6 | Improved |
| add PIC to IQL | IQL · AntMaze average | normalized score: 58.4 -> 78.7 | Improved |
| use one shared hyperparameter setting (delta=1) across all Gym-MuJoCo tasks (ablation) | EPIC · Gym-MuJoCo average | normalized score: 87.8 -> 86.8 | No clear effect |
| use one shared hyperparameter setting (delta=2) across all Gym-MuJoCo tasks (ablation) | EPIC · Gym-MuJoCo average | normalized score: 87.8 -> 87.3 | No clear effect |
| increase PIC strength delta (ablation) | EPIC with critic ensembles · hopper-medium-expert | Qmin / Qstd uncertainty | Improved |
| increase number of peer states K beyond 20 (ablation) | PIC-TD3 and EPIC · Gym-MuJoCo tasks | normalized return | Mixed |
| set PIC strength delta below 1 or above 4 (ablation) | EPIC · hopper-medium | normalized return | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- RORL — RORL: Robust Offline Reinforcement Learning via Conservative Smoothing (NeurIPS 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Policy Regularization with Dataset Constraint for Offline Reinforcement Learning (ICML 2023 · similarity 0.92)
- Adaptive Neighborhood-Constrained Q Learning for Offline Reinforcement Learning (NeurIPS 2025 · similarity 0.91)
- Selective Uncertainty Propagation in Offline RL (AAAI 2025 · similarity 0.91)
- Constrained Latent Action Policies for Model-Based Offline Reinforcement Learning (NeurIPS 2024 · similarity 0.90)
- A Clean Slate for Offline Reinforcement Learning (NeurIPS 2025 · similarity 0.90)
- Adaptive Advantage-Guided Policy Regularization for Offline Reinforcement Learning (ICML 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.
