Research map / Reinforcement learning / Imitation learning
Provably Efficient Learning of Transferable Rewards
Alberto Maria Metelli, Giorgia Ramponi, Alessandro Concetti, Marcello Restelli
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 |
|---|---|---|---|
| adding a meaningful IRL reward-selection criterion (MaxGap, Linear, MaxEnt) instead of random selection from the feasible set (ablation) | TRAVEL · Random MDPs | performance gap ||V*_rE - V^pi*_rE||_2^2 | Improved |
| active sampling (TRAVEL) instead of uniform sampling | IRL with generative model · Chain MDP | performance gap ||V*_rE - V^pi*_rE||_2^2: near-optimal performance in less than half the samples | Improved |
| active sampling (TRAVEL) instead of uniform sampling | IRL with generative model · Chain MDP | performance gap at small sample counts | Mixed |
| IRL (reward transfer) instead of behavioral cloning | policy transfer to target environment · Gridworld with varying obstacle probability | performance gap in target MDP | Improved |
| random reward selection from feasible set (Random-IRL) instead of meaningful criterion (ablation) | TRAVEL · Random MDPs | performance gap | Hurt |
Closest papers by meaning
- Towards the Transferability of Rewards Recovered via Regularized Inverse Reinforcement Learning (NeurIPS 2024 · similarity 0.93)
- Recent Advancements in Inverse Reinforcement Learning (AAAI 2024 · similarity 0.92)
- Is Inverse Reinforcement Learning Harder than Standard Reinforcement Learning? A Theoretical Perspective (ICML 2024 · similarity 0.92)
- Environment Design for Inverse Reinforcement Learning (ICML 2024 · similarity 0.91)
- Offline Inverse RL: New Solution Concepts and Provably Efficient Algorithms (ICML 2024 · similarity 0.90)
- Sub-optimal Experts mitigate Ambiguity in Inverse Reinforcement Learning (NeurIPS 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.
