Research map / Reinforcement learning / Visual and representation RL
Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables
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 off-policy meta-training with probabilistic context variables | PEARL · Six continuous control meta-RL benchmarks | meta-training sample efficiency: 20-100X | Helped |
| Replace the permutation-invariant encoder with an RNN encoder (ablation) | PEARL · Half-Cheetah-Vel | test-task performance | No clear effect |
| Sample RL batches as trajectories instead of decorrelated transitions (ablation) | PEARL · Half-Cheetah-Vel | test-task performance | Hurt |
| Sample context from the entire replay buffer rather than recently collected data (ablation) | PEARL · Continuous control meta-RL | test-task performance | Hurt |
| Reuse the off-policy RL batch as the context batch (ablation) | PEARL · Continuous control meta-RL | test-task performance | Helped |
| Replace the probabilistic context posterior with a deterministic point estimate (ablation) | PEARL · Sparse 2D navigation | task performance | Hurt |
| Train the encoder to maximize actor returns or reconstruct states and rewards instead of recovering the state-action value function (ablation) | PEARL · Continuous control meta-RL | performance | Hurt |
Closest papers by meaning
- Offline Meta-Reinforcement Learning with Online Self-Supervision (ICML 2022 · similarity 0.92)
- Towards an Information Theoretic Framework of Context-Based Offline Meta-Reinforcement Learning (NeurIPS 2024 · similarity 0.92)
- Decoupling Meta-Reinforcement Learning with Gaussian Task Contexts and Skills (AAAI 2024 · similarity 0.91)
- Meta-Reinforcement Learning with Universal Policy Adaptation: Provable Near-Optimality under All-task Optimum Comparator (NeurIPS 2024 · similarity 0.91)
- MetaCARD: Meta-Reinforcement Learning with Task Uncertainty Feedback via Decoupled Context-Aware Reward and Dynamics Components (AAAI 2024 · similarity 0.90)
- Pessimistic Latent Task-aware Optimization for Robust Offline Meta-Reinforcement Learning (NeurIPS 2026 · 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.