Research map / Reinforcement learning / Visual and representation RL
Self-Paced Deep Reinforcement Learning
Pascal Klink, Carlo D'Eramo, Jan R Peters, Joni K. Pajarinen
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 |
|---|---|---|---|
| Use SPDL curriculum generation instead of Default learning | PPO · Point Mass P3D | average final reward: 9.35\pm0.1 vs 2.46\pm0.0 | Improved |
| Use SPDL curriculum generation instead of ALP-GMM | PPO · Point Mass P2D | average final reward: 9.02 \pm 0.4 vs 5.24\pm0.4 | Improved |
| Use SPDL curriculum generation instead of Default learning | PPO · Ant | reward | Improved |
| Use ALP-GMM curriculum generation instead of Default learning | PPO · Ant | reward | No clear effect |
| Use a curriculum rather than sample directly from the target distribution | Ball-Catching agents · Ball-Catching | reward | Improved |
| Initialize the SPDL context distribution (ablation) | Ball-Catching agents · Ball-Catching | average final reward: 47.0\pm2.0 vs 43.3\pm2.0; 53.9\pm0.4 vs 49.3\pm1.4 | No clear effect |
| Use SPDL rather than the other evaluated curricula | TRPO and PPO · Ball-Catching | average final reward | No clear effect |
Closest papers by meaning
- A Probabilistic Interpretation of Self-Paced Learning with Applications to Reinforcement Learning (JMLR 2021 · similarity 0.93)
- TeachMyAgent: a Benchmark for Automatic Curriculum Learning in Deep RL (ICML 2021 · similarity 0.90)
- PORTAL: Automatic Curricula Generation for Multiagent Reinforcement Learning (AAAI 2024 · similarity 0.88)
- Diffusion-based Curriculum Reinforcement Learning (NeurIPS 2024 · similarity 0.88)
- Curriculum Design for Trajectory-Constrained Agent: Compressing Chain-of-Thought Tokens in LLMs (NeurIPS 2025 · similarity 0.87)
- Learning Progress Driven Multi-Agent Curriculum (ICML 2025 · similarity 0.87)
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.
