Research map / Probabilistic and scientific modeling / Molecules and proteins
Neural Bridge Sampling for Evaluating Safety-Critical Autonomous Systems
Aman Sinha, Matthew O'Kelly, Russ Tedrake, John C. Duchi
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 neural bridge sampling instead of naive Monte Carlo | rare-event estimation · Synthetic | relative mean-square error: 1.1821 -> 0.0051 | Improved |
| use neural bridge sampling instead of naive Monte Carlo | rare-event estimation · MountainCar | relative mean-square error: 0.2410 -> 0.0945 | Improved |
| use neural bridge sampling instead of naive Monte Carlo | rare-event estimation · Rocket2 | relative mean-square error: 0.0865 -> 0.0078 | Improved |
| use neural bridge sampling instead of naive Monte Carlo | rare-event estimation · WorldModelRacer | relative mean-square error: 0.9508 -> 0.1218 | Improved |
| add neural warping with normalizing flows to bridge sampling (ablation) | bridge sampling · MountainCar | relative mean-square error: 0.3856 -> 0.0945 | Improved |
| add neural warping with normalizing flows to bridge sampling (ablation) | bridge sampling · Synthetic | relative mean-square error: 0.0514 -> 0.0051 | Improved |
| add neural warping with normalizing flows to bridge sampling (ablation) | bridge sampling · Rocket1 | relative mean-square error: 0.0129 -> 0.0102 | Improved |
| add neural warping with normalizing flows to bridge sampling (ablation) | bridge sampling · AttentionAgentRacer | relative mean-square error: 0.9030 -> 0.2285 | Improved |
| use neural bridge sampling instead of adaptive multilevel splitting | rare-event estimation · MountainCar | relative mean-square error: 0.5424 -> 0.0945 | Improved |
| use adaptive methods (AMS, B) instead of naive Monte Carlo | rare-event estimation · MountainCar | relative mean-square error: 0.2410 -> 0.5424 (AMS) | Hurt |
| use adaptive multilevel splitting instead of naive Monte Carlo | rare-event estimation · Synthetic | relative mean-square error: 1.1821 -> 0.0162 | Improved |
Closest papers by meaning
- Safely Learning Controlled Stochastic Dynamics (NeurIPS 2025 · similarity 0.85)
- A Physics-Informed Machine Learning Framework for Safe and Optimal Control of Autonomous Systems (ICML 2025 · similarity 0.85)
- Position: AI Safety Must Embrace an Antifragile Perspective (ICML 2025 · similarity 0.85)
- DiffScene: Diffusion-Based Safety-Critical Scenario Generation for Autonomous Vehicles (AAAI 2025 · similarity 0.84)
- Safety-Polarized and Prioritized Reinforcement Learning (ICML 2025 · similarity 0.84)
- Safety through feedback in Constrained RL (NeurIPS 2024 · similarity 0.84)
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.
