Meta-Learning 벤치마크
Meta-Learning on ML10
Meta-test success rate
- 2019-10-24 — MAML: Meta-test success rate 36.0
| Rank | Model | Meta-test success rate | Meta-train success rate | Meta-test success rate (zero-shot) | Paper | Code | Year |
|---|---|---|---|---|---|---|---|
| 1 | MAML | 36% | 25% | – | Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning | rlworkgroup/metaworld · farama-foundation/metaworld · avivne/bilinear-transduction · +6 | 2019 |
| 2 | RL^2 | 10% | 50% | – | Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning | rlworkgroup/metaworld · farama-foundation/metaworld · avivne/bilinear-transduction · +6 | 2019 |
| 3 | DnC | 5.4% | – | – | Analyzing Policy Distillation on Multi-Task Learning and Meta-Reinforcement Learning in Meta-World | 2020 | |
| 4 | PEARL | 0% | 42.78% | – | Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning | rlworkgroup/metaworld · farama-foundation/metaworld · avivne/bilinear-transduction · +6 | 2019 |
| 5 | MZ+Recon | – | 97.8% | 25 | Procedural Generalization by Planning with Self-Supervised World Models | 2021 | |
| 6 | MZ | – | 97.6% | 26.5 | Procedural Generalization by Planning with Self-Supervised World Models | 2021 |