Research map / Reinforcement learning / Imitation learning
Bayesian Robust Optimization for Imitation Learning
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 |
|---|---|---|---|
| decrease lambda from 1 (risk-neutral) toward 0, adding CVaR weight (ablation) | BROIL robust objective · Machine replacement | robustness (CVaR) vs expected return | Helped |
| use lambda < 1 instead of optimizing only the mean reward (ablation) | BROIL robust objective · Machine replacement | return distribution tail | Helped |
| use BROIL robust formulation instead of MaxEnt IRL and LPAL | Ambiguous demonstration gridworld · Ambiguous demonstrations | expected return and CVaR over posterior | Helped |
| use baseline regret objective instead of robust objective (ablation) | BROIL · Ambiguous demonstrations | tail performance of baseline regret | Helped |
| increase lambda from 0 in baseline regret BROIL (ablation) | BROIL baseline regret · Ambiguous demonstrations | willingness to take shortcut through red cells | Mixed |
| use a larger Bayesian IRL inverse temperature than beta = 10 (ablation) | BROIL robust policy · Ambiguous demonstrations | match to demonstrator actions | Helped |
Closest papers by meaning
- Simplifying Constraint Inference with Inverse Reinforcement Learning (NeurIPS 2024 · similarity 0.90)
- Rethinking Inverse Reinforcement Learning: from Data Alignment to Task Alignment (NeurIPS 2024 · similarity 0.89)
- Trust Region Reward Optimization and Proximal Inverse Reward Optimization Algorithm (NeurIPS 2025 · similarity 0.89)
- Uncertainty-Based Smooth Policy Regularisation for Reinforcement Learning with Few Demonstrations (NeurIPS 2025 · similarity 0.89)
- Is Inverse Reinforcement Learning Harder than Standard Reinforcement Learning? A Theoretical Perspective (ICML 2024 · similarity 0.89)
- Hybrid Inverse Reinforcement Learning (ICML 2024 · similarity 0.88)
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.