Research map / Reinforcement learning / Visual and representation RL
Data-Efficient 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 |
|---|---|---|---|
| Use HIRO instead of SNN4HRL | HRL on Ant Gather · Ant Gather | average reward: 1.92±0.52 -> 3.02±1.49 | Helped |
| Use HIRO instead of FuN | HRL on Ant Maze · Ant Maze | success rate: 0.0±0.0 -> 0.99±0.01 | Helped |
| Use HIRO instead of FuN transition PG | HRL on Ant Push · Ant Push | success rate: 0.56±0.39 -> 0.92±0.04 | Helped |
| Use HIRO instead of FuN transition PG | HRL on Ant Fall · Ant Fall | success rate: 0.01±0.02 -> 0.66±0.07 | Helped |
| Add random-goal lower-level relabeling (ablation) | HIRO · evaluated tasks | training performance | Mixed |
| Pre-train the lower-level policy for 2M steps and freeze it (ablation) | HIRO · hard navigation tasks | success rate: 2M steps | Hurt |
| Pre-train the lower-level policy for 2M steps and freeze it (ablation) | HIRO · Ant Gather | average reward: 2M steps | Helped |
| Remove off-policy correction (ablation) | HIRO · Ant Push and Ant Fall | success rate | Hurt |
| Remove the hierarchical policy and use a single policy (ablation) | HIRO · evaluated tasks | task performance | Hurt |
| Use learned representations rather than raw observation goals (ablation) | HIRO · evaluated tasks | task performance | Hurt |
Closest papers by meaning
- Overcoming Slow Decision Frequencies in Continuous Control: Model-Based Sequence Reinforcement Learning for Model-Free Control (ICLR 2025 · similarity 0.90)
- Dual RL: Unification and New Methods for Reinforcement and Imitation Learning (ICLR 2024 · similarity 0.90)
- PEAR: Primitive Enabled Adaptive Relabeling for Boosting Hierarchical Reinforcement Learning (ICLR 2025 · similarity 0.90)
- Deep Reinforcement Learning for Robotics: A Survey of Real-World Successes (AAAI 2025 · similarity 0.89)
- Flattening Hierarchies with Policy Bootstrapping (NeurIPS 2025 · similarity 0.89)
- DUO: Diverse, Uncertain, On-Policy Query Generation and Selection for Reinforcement Learning from Human Feedback (AAAI 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.