Research map / Reinforcement learning / Imitation learning
Policy Improvement via Imitation of Multiple Oracles
Ching-An Cheng, Andrey Kolobov, Alekh Agarwal
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 |
|---|---|---|---|
| Use MAMBA with λ=0.9 instead of PG-GAE or AggreVaTeD | Policy optimization with suboptimal oracles · CartPole, DoubleInvertedPendulum, Halfcheetah, Ant | policy return | Improved |
| Use a positive λ rather than λ=0 (ablation) | MAMBA with one oracle · CartPole | policy return | Improved |
| Set λ=0.9 rather than λ=0.5 (ablation) | MAMBA with one oracle · CartPole | policy return | Improved |
| Set λ=0.5 rather than λ=0 or λ=0.9 (ablation) | MAMBA with one oracle · CartPole | policy return | Hurt |
| Increase the oracle set from one to more oracles, including weaker oracles (ablation) | MAMBA with λ=0.9 · DoubleInvertedPendulum, Halfcheetah, Ant | policy optimization performance | Improved |
| Increase the number of oracles (ablation) | MAMBA with λ=0.9 · Higher-dimensional environments | benefit from additional oracles | Mixed |
Closest papers by meaning
- Blending Imitation and Reinforcement Learning for Robust Policy Improvement (ICLR 2024 · similarity 0.92)
- Oracle-Efficient Reinforcement Learning for Max Value Ensembles (NeurIPS 2024 · similarity 0.88)
- Beyond-Expert Performance with Limited Demonstrations: Efficient Imitation Learning with Double Exploration (ICLR 2025 · similarity 0.88)
- Hybrid Inverse Reinforcement Learning (ICML 2024 · similarity 0.88)
- Learning Equilibria from Data: Provably Efficient Multi-Agent Imitation Learning (NeurIPS 2025 · similarity 0.87)
- Agnostic Interactive Imitation Learning: New Theory and Practical Algorithms (ICML 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.
