# From Particles to Perils: SVGD-Based Hazardous Scenario Generation for Autonomous Driving Systems Testing

arXiv: https://arxiv.org/abs/2604.18918 · HTML: https://mlsanity.com/paper/2604.18918/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Adversarial scenarios and safety testing, one of 9 directions in End-to-end autonomous driving.
- Builds on 22 of its 78 references that are in the map.
- First posted on arXiv 20 Apr 2026.

## Placed in

- End-to-end autonomous driving / [Adversarial scenarios and safety testing](https://mlsanity.com/driving/adversarial-scenarios-and-safety-testing/)

## Builds on (most cited first, 22 of 22)

- [Playing Atari with Deep Reinforcement Learning](https://mlsanity.com/paper/1312.5602/) (2013)
- [CARLA: An Open Urban Driving Simulator](https://mlsanity.com/paper/1711.03938/) (2017)
- [Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone](https://mlsanity.com/paper/2404.14219/) (2024)
- [AirSim: High-Fidelity Visual and Physical Simulation for Autonomous Vehicles](https://mlsanity.com/paper/1705.05065/) (2017)
- [Stein Variational Gradient Descent: A General Purpose Bayesian Inference Algorithm](https://mlsanity.com/paper/1608.04471/) (2016)
- [End-to-end Autonomous Driving: Challenges and Frontiers](https://mlsanity.com/paper/2306.16927/) (2023)
- [KING: Generating Safety-Critical Driving Scenarios for Robust Imitation via Kinematics Gradients](https://mlsanity.com/paper/2204.13683/) (2022)
- [Adaptive Stress Testing for Autonomous Vehicles](https://mlsanity.com/paper/1902.01909/) (2019)
- [Large Language Models for UAVs: Current State and Pathways to the Future](https://mlsanity.com/paper/2405.01745/) (2024)
- [Scenario Dreamer: Vectorized Latent Diffusion for Generating Driving Simulation Environments](https://mlsanity.com/paper/2503.22496/) (2025)
- [DriveSceneGen: Generating Diverse and Realistic Driving Scenarios from Scratch](https://mlsanity.com/paper/2309.14685/) (2023)
- [Neural Network Guided Evolutionary Fuzzing for Finding Traffic Violations of Autonomous Vehicles](https://mlsanity.com/paper/2109.06126/) (2021)
- [Efficient and Effective Generation of Test Cases for Pedestrian Detection -- Search-based Software Testing of Baidu Apollo in SVL](https://mlsanity.com/paper/2109.07960/) (2021)
- [Scenario-Based Test Reduction and Prioritization for Multi-Module Autonomous Driving Systems](https://mlsanity.com/paper/2209.01546/) (2022)
- [Multi-Resolution Elevation Mapping and Safe Landing Site Detection with Applications to Planetary Rotorcraft](https://mlsanity.com/paper/2111.06271/) (2021)
- [NeuroStrata: Harnessing Neurosymbolic Paradigms for Improved Design, Testability, and Verifiability of Autonomous CPS](https://mlsanity.com/paper/2502.12267/) (2025)
- [Decictor: Towards Evaluating the Robustness of Decision-Making in Autonomous Driving Systems](https://mlsanity.com/paper/2402.18393/) (2024)
- [Free LSD: Prior-Free Visual Landing Site Detection for Autonomous Planes](https://mlsanity.com/paper/1802.09043/) (2018)
- [Visual Environment Assessment for Safe Autonomous Quadrotor Landing](https://mlsanity.com/paper/2311.10065/) (2023)
- [VisLanding: Monocular 3D Perception for UAV Safe Landing via Depth-Normal Synergy](https://mlsanity.com/paper/2506.14525/) (2025)
- [Local Trajectory Planning For UAV Autonomous Landing](https://mlsanity.com/paper/2111.00495/) (2021)
- [PEACE: Prompt Engineering Automation for CLIPSeg Enhancement for Safe-Landing Zone Segmentation](https://mlsanity.com/paper/2310.00085/) (2023)

## Abstract (arXiv)

Simulation-based testing of autonomous driving systems (ADS) must uncover realistic and diverse failures in dense, heterogeneous traffic. However, existing search-based seeding methods (e.g., genetic algorithms) struggle in high-dimensional spaces, often collapsing to limited modes and missing many failure scenarios. We present PtoP, a framework that combines adaptive random seed generation with Stein Variational Gradient Descent (SVGD) to produce diverse, failure-inducing initial conditions. SVGD balances attraction toward high-risk regions and repulsion among particles, yielding risk-seeking yet well-distributed seeds across multiple failure modes. PtoP is plug-and-play and enhances existing online testing methods (e.g., reinforcement learning--based testers) by providing principled seeds. Evaluation in CARLA on two industry-grade ADS (Apollo, Autoware) and a native end-to-end system shows that PtoP improves safety violation rate (up to 27.68%), scenario diversity (9.6%), and map coverage (16.78%) over baselines.