Neural Network Verification is a Programming Language Challenge
Not on the map. Why?
Abstract (arXiv)
Neural network verification is a new and rapidly developing field of research. So far, the main priority has been establishing efficient verification algorithms and tools, while proper support from the programming language perspective has been considered secondary or unimportant. Yet, there is mounting evidence that insights from the programming language community may make a difference in the future development of this domain. In this paper, we formulate neural network verification challenges as programming language challenges and suggest possible future solutions.
Key facts. Cited by 12 papers in the map (data through 14 Oct 2026). Builds on 65 of its 125 references that are in the map. First posted on arXiv 10 Jan 2025.
Builds on
65 of 125 references.
References · 65
- 2014Explaining and Harnessing Adversarial Examplescited by 7,364
- 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
- 2013Intriguing properties of neural networkscited by 5,155
- 2021A Survey of Quantization Methods for Efficient Neural Network Inferencecited by 815
- 2017Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networkscited by 744
- 2016Safety Verification of Deep Neural Networkscited by 263
- 2017Formal Verification of Piece-Wise Linear Feed-Forward Neural Networkscited by 242
- 2021Beta-CROWN: Efficient Bound Propagation with Per-neuron Split Constraints for Complete and Incomplete Neural Network Robustness Verificationcited by 228
- 2019Towards Stable and Efficient Training of Verifiably Robust Neural Networkscited by 205
- 2020Binary Neural Networks: A Surveycited by 190
- 2018Training Deep Neural Networks with 8-bit Floating Point Numberscited by 164
- 2019Additive Powers-of-Two Quantization: An Efficient Non-uniform Discretization for Neural Networkscited by 148
38 more
- 2020NNV: The Neural Network Verification Tool for Deep Neural Networks and Learning-Enabled Cyber-Physical Systemscited by 128
- 2018Verisig: verifying safety properties of hybrid systems with neural network controllerscited by 84
- 2018Scenic: A Language for Scenario Specification and Scene Generationcited by 83
- 2024Marabou 2.0: A Versatile Formal Analyzer of Neural Networkscited by 81
- 2017Output Reachable Set Estimation and Verification for Multi-Layer Neural Networkscited by 80
- 2017Verifying Properties of Binarized Deep Neural Networkscited by 79
- 2021The Second International Verification of Neural Networks Competition (VNN-COMP 2021): Summary and Resultscited by 71
- 2018TensorFuzz: Debugging Neural Networks with Coverage-Guided Fuzzingcited by 68
- 2021Introduction to Neural Network Verificationcited by 63
- 2021Randomness In Neural Network Training: Characterizing The Impact of Toolingcited by 59
- 2023First Three Years of the International Verification of Neural Networks Competition (VNN-COMP)cited by 58
- 2022Certified Training: Small Boxes are All You Needcited by 52
- 2020An SMT-Based Approach for Verifying Binarized Neural Networkscited by 47
- 2022Deep Learning with Logical Constraintscited by 44
- 2024The Fifth International Verification of Neural Networks Competition (VNN-COMP 2024): Summary and Resultscited by 43
- 2021POLAR: A Polynomial Arithmetic Framework for Verifying Neural-Network Controlled Systemscited by 39
- 2023The Fourth International Verification of Neural Networks Competition (VNN-COMP 2023): Summary and Resultscited by 38
- 2020Exploiting Verified Neural Networks via Floating Point Numerical Errorcited by 35
- 2021Neural Network Robustness as a Verification Property: A Principled Case Studycited by 35
- 2020Efficient Exact Verification of Binarized Neural Networkscited by 33
- 2019JuliaReach: a Toolbox for Set-Based Reachabilitycited by 28
- 2020Scalable Verification of Quantized Neural Networks (Technical Report)cited by 28
- 2021Reachability Is NP-Complete Even for the Simplest Neural Networkscited by 27
- 2019Inductive Logic Programming via Differentiable Deep Neural Logic Networkscited by 25
- 2022Efficient Neural Network Analysis with Sum-of-Infeasibilitiescited by 24
- 2022Open- and Closed-Loop Neural Network Verification using Polynomial Zonotopescited by 24
- 2021DNNV: A Framework for Deep Neural Network Verificationcited by 23
- 2022Neural Network Verification with Proof Productioncited by 22
- 2022Neuro-Symbolic Verification of Deep Neural Networkscited by 21
- 2018Simultaneously Optimizing Weight and Quantizer of Ternary Neural Network using Truncated Gaussian Approximationcited by 20
- 20231-Lipschitz Layers Compared: Memory, Speed, and Certifiable Robustnesscited by 16
- 2024Formally Verifying Deep Reinforcement Learning Controllers with Lyapunov Barrier Certificatescited by 16
- 2023Logic of Differentiable Logics: Towards a Uniform Semantics of DLcited by 15
- 2024Vehicle: Bridging the Embedding Gap in the Verification of Neuro-Symbolic Programscited by 14
- 2023CBMC: The C Bounded Model Checkercited by 12
- 2023Towards Efficient Verification of Quantized Neural Networkscited by 11
- 2022CAISAR: A platform for Characterizing Artificial Intelligence Safety and Robustnesscited by 9
- 2022QEBVerif: Quantization Error Bound Verification of Neural Networkscited by 9
50 most cited of 65 All 65 (JSON)
Built on by
Citing papers per year: 2025: 7; 2026: 5.
12 papers.
- 2025No Soundness in the Real World: On the Challenges of the Verification of Deployed Neural Networkscited by 6
- 2025Proof Minimization in Neural Network Verificationcited by 4
- 2025Floating-Point Neural Network Verification at the Software Levelcited by 2
- 2025Proof-Carrying Neuro-Symbolic Codecited by 1
- 2026VNN-LIB 2.0: Rigorous Foundations for Neural Network Verificationcited by 1
- 2025Neural Network Verification for Gliding Drone Control: A Case Studycited by 0
- 2025The CAISAR Platform: Extending the Reach of Machine Learning Specification and Verificationcited by 0
- 2025Of Good Demons and Bad Angels: Guaranteeing Safe Control under Finite Precisioncited by 0
- 2026Delayed Backdoor Attacks: Exploring the Temporal Dimension as a New Attack Surface in Pre-Trained Modelscited by 0
- 2026Lipschitz-Based Robustness Certification Under Floating-Point Executioncited by 0
- 2026DeepFix: Debugging and Fixing Machine Learning Workflow using Agentic AIcited by 0
- 2026Logic-Gated Time-Shared Feedforward Networks for Alternating Finite Automata: Exact Simulation and Learnabilitycited by 0
- 2026VNN-LIB 2.0: Rigorous Foundations for Neural Network Verification8 May 2026
- 2026Logic-Gated Time-Shared Feedforward Networks for Alternating Finite Automata: Exact Simulation and Learnability20 Mar 2026
- 2026DeepFix: Debugging and Fixing Machine Learning Workflow using Agentic AI14 Mar 2026
- 2026Delayed Backdoor Attacks: Exploring the Temporal Dimension as a New Attack Surface in Pre-Trained Models12 Mar 2026
- 2026Lipschitz-Based Robustness Certification Under Floating-Point Execution6 Mar 2026
- 2025Proof Minimization in Neural Network Verification11 Nov 2025
- 2025Floating-Point Neural Network Verification at the Software Level27 Oct 2025
- 2025Of Good Demons and Bad Angels: Guaranteeing Safe Control under Finite Precision30 Jul 2025
- 2025The CAISAR Platform: Extending the Reach of Machine Learning Specification and Verification10 Jun 2025
- 2025No Soundness in the Real World: On the Challenges of the Verification of Deployed Neural Networks1 Jun 2025
- 2025Neural Network Verification for Gliding Drone Control: A Case Study1 May 2025
- 2025Proof-Carrying Neuro-Symbolic Code16 Apr 2025