Research map / Reinforcement learning / Policy gradient and actor-critic
Learning Control Policies for Stochastic Systems with Reach-Avoid Guarantees
Đorđe Žikelić, Mathias Lechner, Thomas A. Henzinger, Krishnendu Chatterjee
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 RASM level-set reasoning that considers reachability and safety simultaneously instead of the RSM extension | RASM (ours) · 2D system | reach-avoid probability: 83.4% -> 93.3% | Improved |
| Use RASM level-set reasoning that considers reachability and safety simultaneously instead of the RSM extension | RASM (ours) · Inverted pendulum | reach-avoid probability: 47.9% -> 92.1% | Improved |
| Use RASM level-set reasoning that considers reachability and safety simultaneously instead of the RSM extension | RASM (ours) · Collision avoidance | reach-avoid probability: Fail -> 90.4% | Improved |
| Keep policy parameters trainable while learning the RASM instead of freezing the policy (ablation) | 2D system · 2D system | reach-avoid probability: Fail (10 iters.) -> 96.7% (4 iters.) | Improved |
| Keep policy parameters trainable while learning the RASM instead of freezing the policy (ablation) | Collision avoidance · Collision avoidance | reach-avoid probability: Fail (9 iters.) -> 80.9% (3 iters.) | Improved |
| Keep policy parameters trainable while learning the RASM instead of freezing the policy (ablation) | Inverted pendulum · Inverted pendulum | reach-avoid probability: Fail (7 iters.) -> Fail (7 iters.) | No clear effect |
| Use a PPO-pretrained starting policy before joint policy-and-RASM learning | RASM policy learning · Inverted pendulum | reach-avoid probability | Mixed |
Closest papers by meaning
- Safely Learning Controlled Stochastic Dynamics (NeurIPS 2025 · similarity 0.87)
- Absolute State-wise Constrained Policy Optimization: High-Probability State-wise Constraints Satisfaction (NeurIPS 2026 · similarity 0.87)
- Neural Continuous-Time Supermartingale Certificates (AAAI 2025 · similarity 0.86)
- Regret-Free Reinforcement Learning for Temporal Logic Specifications (ICML 2025 · similarity 0.86)
- Verified Safe Reinforcement Learning for Neural Network Dynamic Models (NeurIPS 2024 · similarity 0.86)
- Learning Optimal Deterministic Policies with Stochastic Policy Gradients (ICML 2024 · similarity 0.86)
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.
