mlsanityData to 14 Oct 2026

DeepXplore: Automated Whitebox Testing of Deep Learning Systems

arXiv 1705.06640 ↗First posted 18 May 2017cs.LG cs.CR cs.SEcited by 311

Not on the map. Why?

Abstract (arXiv)

Deep learning (DL) systems are increasingly deployed in safety- and security-critical domains including self-driving cars and malware detection, where the correctness and predictability of a system's behavior for corner case inputs are of great importance. Existing DL testing depends heavily on manually labeled data and therefore often fails to expose erroneous behaviors for rare inputs. We design, implement, and evaluate DeepXplore, the first whitebox framework for systematically testing real-world DL systems. First, we introduce neuron coverage for systematically measuring the parts of a DL system exercised by test inputs. Next, we leverage multiple DL systems with similar functionality as cross-referencing oracles to avoid manual checking. Finally, we demonstrate how finding inputs for DL systems that both trigger many differential behaviors and achieve high neuron coverage can be represented as a joint optimization problem and solved efficiently using gradient-based search techniques. DeepXplore efficiently finds thousands of incorrect corner case behaviors (e.g., self-driving cars crashing into guard rails and malware masquerading as benign software) in state-of-the-art DL models with thousands of neurons trained on five popular datasets including ImageNet and Udacity self-driving challenge data. For all tested DL models, on average, DeepXplore generated one test input demonstrating incorrect behavior within one second while running only on a commodity laptop. We further show that the test inputs generated by DeepXplore can also be used to retrain the corresponding DL model to improve the model's accuracy by up to 3%.

Key facts. Cited by 311 papers in the map (data through 14 Oct 2026). First posted on arXiv 18 May 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 71; 2022: 57; 2023: 63; 2024: 49; 2025: 40; 2026: 31.

311 papers.

Sort:

By area:

  1. 2023Are aligned neural networks adversarially aligned?cited by 315
  2. 2021Trustworthy AI: From Principles to Practicescited by 207
  3. 2021Invisible for both Camera and LiDAR: Security of Multi-Sensor Fusion based Perception in Autonomous Driving Under Physical-World Attackscited by 101
  4. 2021Explainability of deep vision-based autonomous driving systems: Review and challengescited by 80
  5. 2023How Deep Learning Sees the World: A Survey on Adversarial Attacks & Defensescited by 66
  6. 2022Physical Attack on Monocular Depth Estimation with Optimal Adversarial Patchescited by 57
  7. 2023BiasAsker: Measuring the Bias in Conversational AI Systemcited by 53
  8. 2023Evaluating Superhuman Models with Consistency Checkscited by 47
  9. 2021Copy, Right? A Testing Framework for Copyright Protection of Deep Learning Modelscited by 46
  10. 2021DeepPayload: Black-box Backdoor Attack on Deep Learning Models through Neural Payload Injectioncited by 44
  11. 2021A Comprehensive Evaluation Framework for Deep Model Robustnesscited by 42
  12. 2021Guidance on the Assurance of Machine Learning in Autonomous Systems (AMLAS)cited by 42
38 more
  1. 2021Provable Repair of Deep Neural Networkscited by 42
  2. 2021NeuronFair: Interpretable White-Box Fairness Testing through Biased Neuron Identificationcited by 42
  3. 2021Robust Machine Learning Systems: Challenges, Current Trends, Perspectives, and the Road Aheadcited by 41
  4. 2021How to Certify Machine Learning Based Safety-critical Systems? A Systematic Literature Reviewcited by 39
  5. 2021Mind the Gap! A Study on the Transferability of Virtual vs Physical-world Testing of Autonomous Driving Systemscited by 37
  6. 2023WhiteFox: White-Box Compiler Fuzzing Empowered by Large Language Modelscited by 37
  7. 2021Multi-Label Classification Neural Networks with Hard Logical Constraintscited by 36
  8. 2026ReliabilityBench: Evaluating LLM Agent Reliability Under Production-Like Stress Conditionscited by 36
  9. 2021Description of Corner Cases in Automated Driving: Goals and Challengescited by 35
  10. 2021ModelDiff: Testing-Based DNN Similarity Comparison for Model Reuse Detectioncited by 33
  11. 2023Neuron Structure Modeling for Generalizable Remote Physiological Measurementcited by 31
  12. 2021DeepHyperion: Exploring the Feature Space of Deep Learning-Based Systems through Illumination Searchcited by 29
  13. 2023Neuron Activation Coverage: Rethinking Out-of-distribution Detection and Generalizationcited by 28
  14. 2022Simple Techniques Work Surprisingly Well for Neural Network Test Prioritization and Active Learning (Replicability Study)cited by 27
  15. 2021Taxonomy of Machine Learning Safety: A Survey and Primercited by 25
  16. 2021Can Offline Testing of Deep Neural Networks Replace Their Online Testing?cited by 24
  17. 2024Adversarial Attacks of Vision Tasks in the Past 10 Years: A Surveycited by 23
  18. 2021Distribution-Aware Testing of Neural Networks Using Generative Modelscited by 22
  19. 2022SoK: On the Semantic AI Security in Autonomous Drivingcited by 22
  20. 2023MTTM: Metamorphic Testing for Textual Content Moderation Softwarecited by 22
  21. 2024ImageNet-D: Benchmarking Neural Network Robustness on Diffusion Synthetic Objectcited by 22
  22. 2021An Empirical Study on Deployment Faults of Deep Learning Based Mobile Applicationscited by 20
  23. 2021Neuron Coverage-Guided Domain Generalizationcited by 19
  24. 2021Robust Semantic Segmentation with Superpixel-Mixcited by 19
  25. 2021Efficient Out-of-Distribution Detection Using Latent Space of β-VAE for Cyber-Physical Systemscited by 19
  26. 2021Black-Box Testing of Deep Neural Networks Through Test Case Diversitycited by 18
  27. 2022NPC: Neuron Path Coverage via Characterizing Decision Logic of Deep Neural Networkscited by 18
  28. 2022When and Why Test Generators for Deep Learning Produce Invalid Inputs: an Empirical Studycited by 18
  29. 2024Sample-Efficient Human Evaluation of Large Language Models via Maximum Discrepancy Competitioncited by 18
  30. 2021DeepGuard: A Framework for Safeguarding Autonomous Driving Systems from Inconsistent Behaviorcited by 17
  31. 2022DeepPerform: An Efficient Approach for Performance Testing of Resource-Constrained Neural Networkscited by 17
  32. 2023Convolutional Visual Prompt for Robust Visual Perceptioncited by 17
  33. 2024AcTracer: Active Testing of Large Language Model via Multi-Stage Samplingcited by 17
  34. 2022LLMEffiChecker: Understanding and Testing Efficiency Degradation of Large Language Modelscited by 16
  35. 2024New Job, New Gender? Measuring the Social Bias in Image Generation Modelscited by 16
  36. 2022AEON: A Method for Automatic Evaluation of NLP Test Casescited by 15
  37. 2022Bugs in Machine Learning-based Systems: A Faultload Benchmarkcited by 15
  38. 2024Are LLMs Correctly Integrated into Software Systems?cited by 15

50 most cited of 311 All 311 (JSON)