Research map / Reinforcement learning / Policy gradient and actor-critic
Off-Policy Safe Reinforcement Learning with Cost-Constrained Optimistic Exploration
Guopeng Li, Matthijs T. J. Spaan, Julian F. P. Kooij
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 |
|---|---|---|---|
| cost-constrained exploration (COX) instead of ORAC-style exploration (ablation) | TQC-based COX-Q · SMARTS Overtaking | timeouts at test: 887 -> 0 | Improved |
| COX exploration instead of ORAC-style exploration (ablation) | TQC-based COX-Q · SMARTS Overtaking | unsafe events in data collection: 3215 -> 1665 | Improved |
| COX exploration instead of ORAC-style exploration (ablation) | TQC-based COX-Q · SMARTS Intersection | collisions at test: 18 -> 12 | Improved |
| COX exploration instead of ORAC-style exploration (ablation) | TQC-based COX-Q · SMARTS Overtaking | collisions at test: 97 -> 99 | No clear effect |
| COX-Q vs on-policy CPPOPID | COX-Q · SMARTS Intersection | collisions at test: 183 -> 12 | Improved |
| COX-Q vs CAL on data-collection unsafe events | COX-Q · SMARTS Intersection | unsafe events in data collection: 739 -> 1123 | Hurt |
| add TQC truncated quantile critics (ablation) | COX-Q · Safe Velocity and Safe Navigation | return | Improved |
| add cost-constrained exploration on top of TQC (ablation) | COX-Q · Safe Velocity | training cost | Improved |
| add COX or ORAC exploration on top of TQC (ablation) | COX-Q · Safe Navigation | return and cost | No clear effect |
Closest papers by meaning
- Sampling-Based Safe Reinforcement Learning (NeurIPS 2026 · similarity 0.91)
- Safe Reinforcement Learning with Instantaneous Constraints: The Role of Aggressive Exploration (AAAI 2024 · similarity 0.90)
- Latent Safety-Constrained Policy Approach for Safe Offline Reinforcement Learning (ICLR 2025 · similarity 0.90)
- Safe Reinforcement Learning using Finite-Horizon Gradient-based Estimation (ICML 2024 · similarity 0.90)
- Online Optimization for Offline Safe Reinforcement Learning (NeurIPS 2025 · similarity 0.90)
- QPRL : Learning Optimal Policies with Quasi-Potential Functions for Asymmetric Traversal (ICML 2025 · similarity 0.90)
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.
