mlsanityData to 7 Oct 2026

Adversarial scenarios and safety testing

141 papers0 this weekMedium confidence in this grouping: mostly one topic, with blurry edgesMedium confidence in this grouping: mostly one topic, with blurry edges · About

Keywords: adversarial, driving, scenarios, autonomous, testing

Key facts. Adversarial scenarios and safety testing: 141 papers on arXiv since 2021 (data through 7 Oct 2026). 24 of its 141 mapped papers (17%) cite “Generating Useful Accident-Prone Driving Scenarios via a Learned Traffic Prior” (2021, arXiv 2112.05077). 17 of its 141 mapped papers (12%) cite “CAT: Closed-loop Adversarial Training for Safe End-to-End Driving” (2023, arXiv 2310.12432). 4 new papers 17 Sep–7 Oct.

What’s new

17 Sep – 7 Oct 2026

  1. Earlier · 4
  2. Human Behavior-Informed Crash Scenario Generation with Real-World Crash Priors for Autonomous Vehicle Safety Evaluation
  3. TrafficSignBench: Rule-Centric Closed-Loop Evaluation of Traffic-Sign Compliance in Autonomous Driving also fits: Imitation-learning end-to-end planners
  4. ExceptionDrive: A Planning-Oriented Counterfactual Corner-Case Benchmark for Autonomous Driving also fits: End-to-end system architecture
  5. NavSafe-∞: Benchmarking Closed-Loop Driving Safety in Photorealistic Environments

Foundations

Most cited by papers here

  1. Generating Useful Accident-Prone Driving Scenarios via a Learned Traffic Prior2021cited by 24 here
  2. CAT: Closed-loop Adversarial Training for Safe End-to-End Driving2023cited by 17 here
  3. SafeBench: A Benchmarking Platform for Safety Evaluation of Autonomous Vehicles2022cited by 10 here
  4. Corner Case Generation and Analysis for Safety Assessment of Autonomous Vehicles2021cited by 8 here
  5. Invisible for both Camera and LiDAR: Security of Multi-Sensor Fusion based Perception in Autonomous Driving Under Physical-World Attacks2021cited by 8 here

Lineage

The papers this direction builds on, oldest first.

Each paper on the line cites the one before.

  1. 2021End-to-end Interpretable Neural Motion Plannercited by 6 here
  2. 2021Generating Useful Accident-Prone Driving Scenarios via a Learned Traffic Priorcited by 24 here
  3. 2022KING: Generating Safety-Critical Driving Scenarios for Robust Imitation via Kinematics Gradientscited by 23 here
  4. 2022TransFuser: Imitation with Transformer-Based Sensor Fusion for Autonomous Drivingcited by 9 here
  5. 2023End-to-end Autonomous Driving: Challenges and Frontierscited by 15 here
  6. 2023DriveGPT4: Interpretable End-to-end Autonomous Driving via Large Language Modelcited by 10 here
  7. 2023LMDrive: Closed-Loop End-to-End Driving with Large Language Modelscited by 7 here
  8. 2023DriveLM: Driving with Graph Visual Question Answeringcited by 10 here
  9. 2024NAVSIM: Data-Driven Non-Reactive Autonomous Vehicle Simulation and Benchmarkingcited by 7 here
15 more it builds on, not in this chain
  1. 2021Corner Case Generation and Analysis for Safety Assessment of Autonomous Vehiclescited by 8 here
  2. 2021Adversarial Driving: Attacking End-to-End Autonomous Drivingcited by 6 here
  3. 2021Multi-Modal Fusion Transformer for End-to-End Autonomous Drivingcited by 6 here
  4. 2021Invisible for both Camera and LiDAR: Security of Multi-Sensor Fusion based Perception in Autonomous Driving Under Physical-World Attackscited by 8 here
  5. 2021NuPlan: A closed-loop ML-based planning benchmark for autonomous vehiclescited by 14 here
  6. 2021MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learningcited by 17 here
  7. 2022Trajectory-guided Control Prediction for End-to-end Autonomous Driving: A Simple yet Strong Baselinecited by 7 here
  8. 2022SafeBench: A Benchmarking Platform for Safety Evaluation of Autonomous Vehiclescited by 10 here
  9. 2023VAD: Vectorized Scene Representation for Efficient Autonomous Drivingcited by 16 here
  10. 2023Waymax: An Accelerated, Data-Driven Simulator for Large-Scale Autonomous Driving Researchcited by 7 here
  11. 2023CAT: Closed-loop Adversarial Training for Safe End-to-End Drivingcited by 17 here
  12. 2023Dolphins: Multimodal Language Model for Drivingcited by 9 here
  13. 2024ChatScene: Knowledge-Enabled Safety-Critical Scenario Generation for Autonomous Vehiclescited by 10 here
  14. 2024Bench2Drive: Towards Multi-Ability Benchmarking of Closed-Loop End-To-End Autonomous Drivingcited by 12 here
  15. 2025SimLingo: Vision-Only Closed-Loop Autonomous Driving with Language-Action Alignmentcited by 7 here

The 24 most cited of 290.

Browse

30 of 141, ranked 6–35

  1. 6Attack End-to-End Autonomous Driving through Module-Wise Noise20246
  2. 7Visual Adversarial Attack on Vision-Language Models for Autonomous Driving20246
  3. 8Adversarial Driving: Attacking End-to-End Autonomous Driving20216
  4. 9Targeted Attack on Deep RL-based Autonomous Driving with Learned Visual Patterns20216
  5. 10Black-Box Adversarial Attack on Vision Language Models for Autonomous Driving20255
  6. 11Adversarial Safety-Critical Scenario Generation using Naturalistic Human Driving Priors20245
  7. 12(Re)2H2O: Autonomous Driving Scenario Generation via Reversely Regularized Hybrid Offline-and-Online Reinforcement Learning20235
  8. 13CaDRE: Controllable and Diverse Generation of Safety-Critical Driving Scenarios using Real-World Trajectories20244
  9. 14Enhancing Autonomous Vehicle Training with Language Model Integration and Critical Scenario Generation20244
  10. 15DR2L: Surfacing Corner Cases to Robustify Autonomous Driving via Domain Randomization Reinforcement Learning20214
All 30
  1. 16NeuroNCAP: Photorealistic Closed-loop Safety Testing for Autonomous Driving20243
  2. 17Module-wise Adaptive Adversarial Training for End-to-end Autonomous Driving20243
  3. 18SEAL: Towards Safe Autonomous Driving via Skill-Enabled Adversary Learning for Closed-Loop Scenario Generation20243
  4. 19Evaluating Adversarial Attacks on Driving Safety in Vision-Based Autonomous Vehicles20213
  5. 20Neural Network Guided Evolutionary Fuzzing for Finding Traffic Violations of Autonomous Vehicles20213
  6. 21UniAda: Universal Adaptive Multi-objective Adversarial Attack for End-to-End Autonomous Driving Systems20262
  7. 22Challenger: Affordable Adversarial Driving Video Generation20252
  8. 23Pseudo-Simulation for Autonomous Driving20252
  9. 24Sharpening the Spear: Adaptive Expert-Guided Adversarial Attack Against DRL-based Autonomous Driving Policies20252
  10. 25Adversarial Generation and Collaborative Evolution of Safety-Critical Scenarios for Autonomous Vehicles20252
  11. 26Are NeRFs ready for autonomous driving? Towards closing the real-to-simulation gap20242
  12. 27Stackelberg Driver Model for Continual Policy Improvement in Scenario-Based Closed-Loop Autonomous Driving20232
  13. 28Generating Transferable Adversarial Simulation Scenarios for Self-Driving via Neural Rendering20232
  14. 29On Robustness of Lane Detection Models to Physical-World Adversarial Attacks in Autonomous Driving20212
  15. 30From Words to Collisions: LLM-Guided Evaluation and Adversarial Generation of Safety-Critical Driving Scenarios20251
  16. 31Natural Reflection Backdoor Attack on Vision Language Model for Autonomous Driving20251
  17. 32REACT: Runtime-Enabled Active Collision-avoidance Technique for Autonomous Driving20251
  18. 33PhysPatch: A Physically Realizable and Transferable Adversarial Patch Attack for Multimodal Large Language Models-based Autonomous Driving Systems20251
  19. 34Robust Driving Control for Autonomous Vehicles: An Intelligent General-sum Constrained Adversarial Reinforcement Learning Approach20251
  20. 35Quantitative Representation of Scenario Difficulty for Autonomous Driving Based on Adversarial Policy Search20241