Research map / Reinforcement learning / Policy gradient and actor-critic
A Reward-Free Viewpoint on Multi-Objective Reinforcement Learning
Ying-Tu Chen, Wei Hung, Bing-Shu Wu, Zhang-Wei Hong, Ping-Chun Hsieh
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 |
|---|---|---|---|
| add preference-guided z sampling (PG-Explore) instead of standard normal z (ablation) | MORL-FB · MO-Gymnasium Ant3d | utility and hypervolume | Improved |
| add auxiliary Q loss on observed reward vectors (ablation) | MORL-FB · MO-Gymnasium Ant3d | utility and hypervolume | Improved |
| use MORL-FB instead of vanilla FB | forward-backward RFRL · MO-Gymnasium continuous control | utility, hypervolume, episodic dominance | Improved |
| use MORL-FB instead of PD-MORL and Q-Pensieve | MORL baselines · MO-Gymnasium continuous control | utility, hypervolume, ED | Improved |
| train only on one-hot and uniform preferences | MORL-FB · MO-Gymnasium continuous control, limited training preferences | utility, hypervolume, ED | No clear effect |
| transfer from Hopper2d to Hopper3d/4d zero-shot with MORL-FB instead of vanilla FB | forward-backward RFRL · Hopper cross-objective transfer | return | Improved |
Closest papers by meaning
- On Generalization Across Environments In Multi-Objective Reinforcement Learning (ICLR 2025 · similarity 0.92)
- An Analytical Study of Utility Functions in Multi-Objective Reinforcement Learning (NeurIPS 2024 · similarity 0.92)
- Preference Conditioned Multi-Objective Reinforcement Learning: Decomposed, Diversity-Driven Policy Optimization (NeurIPS 2026 · similarity 0.91)
- Multi-objective Linear Reinforcement Learning with Lexicographic Rewards (ICML 2025 · similarity 0.91)
- Pareto Set Learning for Multi-Objective Reinforcement Learning (AAAI 2025 · similarity 0.90)
- Preference Controllable Reinforcement Learning with Advanced Multi-Objective Optimization (ICML 2025 · 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.
