Research map / Reinforcement learning / Imitation learning
Inverse Reward Design
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 |
|---|---|---|---|
| Infer true reward via IRD posterior (MaxEnt-Z or Sample-Z) and plan risk-averse instead of planning with the literal proxy reward | Lavaland navigation agent · Lavaland negative side effects and reward hacking (proof-of-concept) | Fraction of trajectories encountering lava | Helped |
| Use IRD posterior with risk-averse planning instead of literal proxy in latent-reward setting | Lavaland latent-reward agent with raw observations or classifier features · Lavaland challenge domain | Fraction of trajectories encountering lava | Helped |
| Switch from raw-observation proxy to classifier-feature proxy | Lavaland latent-reward setting · Lavaland challenge domain | Fraction of trajectories encountering lava | Hurt |
| Risk-averse planning over the IRD posterior is limited: it also avoids good terrain like the pot of gold | IRD agent · Lavaland | qualitative (not measured) | Mixed |
Closest papers by meaning
- Recent Advancements in Inverse Reinforcement Learning (AAAI 2024 · similarity 0.88)
- Offline Inverse RL: New Solution Concepts and Provably Efficient Algorithms (ICML 2024 · similarity 0.87)
- Is Inverse Reinforcement Learning Harder than Standard Reinforcement Learning? A Theoretical Perspective (ICML 2024 · similarity 0.87)
- A Unified Linear Programming Framework for Offline Reward Learning from Human Demonstrations and Feedback (ICML 2024 · similarity 0.86)
- A Geometric Perspective on Reward Function Updates in Inverse Reinforcement Learning (NeurIPS 2026 · similarity 0.86)
- Partial Identifiability in Inverse Reinforcement Learning for Agents with Non-Exponential Discounting (AAAI 2025 · similarity 0.86)
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.