Research map / Reinforcement learning / Visual and representation RL
Improving Generalization in Meta-RL with Imaginary Tasks from Latent Dynamics Mixture
Suyoung Lee, Sae-Young Chung
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 latent dynamics mixture imaginary tasks to RL^2/variBAD-style agent | LDM · Gridworld M_test | test goals reached: 19 of 27 | Improved |
| apply input dropout to reward decoder (p_drop 0.0 to higher rates) (ablation) | LDM · Gridworld | test return | Improved |
| increase decoder dropout rate to 0.9 (ablation) | LDM · Gridworld | test return | Hurt |
| no decoder input dropout (p_drop=0.0) (ablation) | LDM · Gridworld | test return vs RL^2 | No clear effect |
| apply dropout to policy encoder state input (ablation) | RL^2 · Gridworld | training stability | Hurt |
| share one encoder for policy and VAE with decoder dropout (ablation) | LDM · Gridworld | policy training stability | Hurt |
| increase extrapolation level beta from 1.0 to larger values (ablation) | LDM · Gridworld-extrapolation | return on M_test2 vs M_test1 | Mixed |
| use beta=2 instead of 1 (ablation) | LDM · Ant-goal | test return | Improved |
| vary number of mixture workers (2,3,4,5 of 16) (ablation) | LDM · Half-cheetah-velocity | test return | Mixed |
| mix observations and rewards in observation space (Mixreg) instead of latent space | RL^2 · Gridworld M_test | test return | Mixed |
| Mixreg on MuJoCo tasks | RL^2 · MuJoCo M_eval | test return | Hurt |
Closest papers by meaning
- DynaMITE-RL: A Dynamic Model for Improved Temporal Meta-Reinforcement Learning (NeurIPS 2024 · similarity 0.88)
- MAMBA: an Effective World Model Approach for Meta-Reinforcement Learning (ICLR 2024 · similarity 0.88)
- AMAGO-2: Breaking the Multi-Task Barrier in Meta-Reinforcement Learning with Transformers (NeurIPS 2024 · similarity 0.88)
- Meta-Reinforcement Learning via Exploratory Task Clustering (AAAI 2024 · similarity 0.87)
- Meta-DT: Offline Meta-RL as Conditional Sequence Modeling with World Model Disentanglement (NeurIPS 2024 · similarity 0.87)
- Meta-Reinforcement Learning with Universal Policy Adaptation: Provable Near-Optimality under All-task Optimum Comparator (NeurIPS 2024 · similarity 0.87)
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.
