SoSyM Journal Submission Artifacts; Updates for other upcoming submissions on NNV

This primarily is a release for a journal submission to SoSym in December 2025, based on an extension of a 2025 Formalise paper "Robustness Verification of Video Classification Neural Networks": https://doi.org/10.1109/FormaliSE66629.2025.00009

What's Changed

• Add dockerfile and instructions by @mldiego in #275
• Integrate CP method with default python version by @mldiego in #277
• Add example conformal verification by @mldiego in #278
• Fixing test suite by @mldiego in #279
• Fix CP for tests by @mldiego in #282
• Merging latest changes before copilot tests by @mldiego in #283
• Add comprehensive unit and integration tests for engine/nn layers and functions by @ttj in #281
• Testing updates: soundness and regressions; minor bug fixes by @ttj in #286
• Changes for MNIST MLP experiment by @m-usama-z in #287
• added load_images_MNIST() to utils by @m-usama-z in #289
• SoSym Extension by @sammsaski in #291

New Contributors

• @Copilot made their first contribution in #281
• @ttj made their first contribution in #286
• @m-usama-z made their first contribution in #287

Full Changelog: vnncomp2025...sosym2025

VNNCOMP 2025

VNNCOMP 2025

Official submission by NNV Team to VNNCOMP2025

Benchmarks considered:

• Regular track:
  • ALL
• Extended track:
  • ml4acopf
  • relusplitter
  • vggnet16

Code: https://github.com/verivital/nnv/tree/master/code/nnv/examples/Submission/VNN_COMP2025

What's Changed

• SPIE tutorial by @mldiego in #256
• CP reachability by @Navidhashemicodes
• VNNCOMP 2025 by @mldiego in #274

Full Changelog: formalise2025...vnncomp2025

FormaliSE 2025

Robustness Verification of Video Classification Neural Networks

Published at the International Conference on Formal Methods in Software Engineering, 2025

Code available at: Submission/FORMALISE2025

ICAIF 2024

FairNNV: The Neural Network Verification Tool For Certifying Fairness

Published at the 5th ACM International Conference on AI in Finance

Code available at: Submission/ICAIF24

vnncomp2024

NNV version for the participation in the Verification of Neural Network competition, 2024.
Submission folder: examples/Submission/VNN_COMP2024

Competition Information

5th International Verification of Neural Networks Competition (VNN-COMP'24)

cav2023

CAV 2023 Artifact Evaluation: NNV 2.0: The Neural Network Verification Tool

This is the release corresponding to results in the CAV'23 paper: NNV 2.0: The Neural Network Verification Tool

CodeOcean Capsule: https://codeocean.com/capsule/6689683/

The examples used in the paper are available in code/nnv/examples/NNV2.0/Submission/CAV2023
https://github.com/verivital/nnv/tree/master/code/nnv/examples/NNV2.0/Submission/CAV2023

HSCC2023

Verification of Recurrent Neural Networks (RNN), HSCC2023

FORMATS 2022 Artifact Evaluation: Reachability Analysis of a General Class of Neural Ordinary Differential Equations (ODEs)

This release contains the neural ordinary differential equation (ODE) reachability analysis code, corresponding to the paper "Reachability Analysis of a General Class of Neural Ordinary Differential Equations" by Diego Manzanas Lopez, Patrick Musau, Nathaniel Hamilton and Taylor T Johnson appearing at 20th International Conference on Formal Modeling and Analysis of Timed Systems (FORMATS 2022).

The examples, case studies, etc., used in the paper are available in code/nnv/examples/Submission/FORMATS2022:

Tagged link: https://github.com/verivital/nnv/tree/formats2022/code/nnv/examples/Submission/FORMATS2022

Main link: https://github.com/verivital/nnv/tree/master/code/nnv/examples/Submission/FORMATS2022

FAOC 2021 Examples and Case Studies (Journal Extension of FM'19 Star Set Paper)

This release contains the improved star set reachability methods implemented in NNV, used in the Formal Aspects of Computing (FAOC) journal special issue, extending the earlier Formal Methods (FM'19) star set paper.

The examples, case studies, etc., used in the paper are available in code/nnv/examples/Submission/FM2019_Journal:

https://github.com/verivital/nnv/tree/faoc2021/code/nnv/examples/Submission/FM2019_Journal

CAV 2021 Artifact Evaluation: Robustness Verification of Semantic Segmentation Neural Networks using Relaxed Reachability

This is the release corresponding to results in the CAV'21 paper: Robustness Verification of Semantic Segmentation Neural Networks using Relaxed Reachability