Research map / Reinforcement learning / RL from feedback
Meta-Reward-Net: Implicitly Differentiable Reward Learning for Preference-based Reinforcement 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 |
|---|---|---|---|
| add bi-level reward learning (MRN) on top of PEBBLE | PEBBLE · Walker | Q-function MSE: 0.12 ± 0.04 -> 0.10 ± 0.02 | Helped |
| add bi-level reward learning (MRN) on top of PEBBLE | PEBBLE · Door Open | Q-function MSE: 0.52 ± 0.05 -> 0.48 ± 0.06 | Helped |
| add SURF to PEBBLE | PEBBLE · Door Open | Q-function MSE: 0.52 ± 0.05 -> 0.60 ± 0.32 | Hurt |
| add SURF to PEBBLE | PEBBLE · Walker | Q-function MSE: 0.12 ± 0.04 -> 0.11 ± 0.04 | No clear effect |
| increase number of preference labels from 100 to 2000 (ablation) | MRN vs PEBBLE vs SURF · Walker and Door Open | return / success rate | Mixed |
| use bi-level optimization (MRN) with few preference labels | PEBBLE · Meta-world and DMControl | success rate / return | Helped |
| add SURF to PEBBLE | PEBBLE · DMControl | return | No clear effect |
Closest papers by meaning
- O-MAPL: Offline Multi-agent Preference Learning (ICML 2025 · similarity 0.90)
- Provable Reward-Agnostic Preference-Based Reinforcement Learning (ICLR 2024 · similarity 0.90)
- Decoding Global Preferences: Temporal and Cooperative Dependency Modeling in Multi-Agent Preference-Based Reinforcement Learning (AAAI 2024 · similarity 0.90)
- RIME: Robust Preference-based Reinforcement Learning with Noisy Preferences (ICML 2024 · similarity 0.89)
- Policy-labeled Preference Learning: Is Preference Enough for RLHF? (ICML 2025 · similarity 0.89)
- Reward Model Learning vs. Direct Policy Optimization: A Comparative Analysis of Learning from Human Preferences (ICML 2024 · similarity 0.89)
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.