Research map / Reinforcement learning / Visual and representation RL
An Efficient Approach to Model-Based Hierarchical 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 |
|---|---|---|---|
| share transition experience across subtasks via fragments and component factorization (CSRL) vs R-MAXQ | CSRL · Robot pickup and place (7200 states) | episodes to converge: fewer than 100 episodes vs not converged after 1000 | Helped |
| CSRL without known DBN structure vs Factored R-MAX with DBN given | CSRL · Robot pickup and place (7200 states) | average reward | Helped |
| directed (R-MAX style) exploration vs epsilon-greedy | Factored methods · Robot pickup and place (7200 states) | reward | Helped |
| SMDP task selection with simulation using root transition vs SMDP-RMAX / random task selection | CSRL · Pickup and place, 2 objects (518400 states) | average and accumulated reward | Helped |
| SMDP-RMAX task selection vs random task selection | task selection · Pickup and place, 2 objects | reward | No clear effect |
| SMDP-based task selection vs fixed order task selection (Multi-CSRL vs Multi-FRMAX) | Multi-CSRL · Household robot (1638400 states) | average reward: about -1200 to more than 100 at ~25th episode | Helped |
| Multi-CSRL simulation overhead vs Multi-FRMAX | Multi-CSRL · Household robot | running time for 100 episodes (s): 610.8 vs 616.7 | No clear effect |
Closest papers by meaning
- Data-Efficient Hierarchical Reinforcement Learning (NeurIPS 2018 · similarity 0.91)
- Offline Hierarchical Reinforcement Learning via Inverse Optimization (ICLR 2025 · similarity 0.90)
- Globally Optimal Hierarchical Reinforcement Learning for Linearly-Solvable Markov Decision Processes (AAAI 2022 · similarity 0.89)
- Hierarchical Imitation and Reinforcement Learning (ICML 2018 · similarity 0.89)
- Overcoming Slow Decision Frequencies in Continuous Control: Model-Based Sequence Reinforcement Learning for Model-Free Control (ICLR 2025 · similarity 0.89)
- Reward-Consistent Dynamics Models are Strongly Generalizable for Offline Reinforcement Learning (ICLR 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.