Neural Verification

Automated verification of neural networks, machine learned and standard programs.

Projects

pyNeVer

pyNeVer

API Beta

A Python library for learning and verification of neural networks and other machine learning models.

CoCoNet

CoCoNet

Tool Alpha

A tool for construction and conversion of neural networks across different standards.

Examples
User Guide

NeVer 2

NeVer 2

Tool Alpha

A tool for learning and verification of neural networks.

Examples
Feed-Forward NN Convolutional NN

sCan

SCAn

Tool In development

A statistical code analyzer.

VNN-LIB

VNN-LIB

Initiative

An international initiative whose aim is to encourage collaboration and facilitate research and development in Verification of Neural Networks (VNN).

About

This is the home page of neuralverification.org, an initiative of researchers working at the Department of Informatics, Bioengineering, Robotics and Systems Engineering of the University of Genoa.

The aim of the initiative is to develop tools that can verify neural networks and other machine learned programs as well as standard, i.e., manually written, code.

Showcase



Publications

  • 2024

    • S. Demarchi, D. Guidotti, L. Pulina and A. Tacchella

      NeVer2: Learning and Verification of Neural Networks Article

      Soft Computing - accepted for publication

    • S. Demarchi, A. Gimelli and A. Tacchella

      Improving Abstract Propagation for Verification of Neural Networks Conference

      International ECMS Conference on Modeling and Simulation

  • 2023

    • S. Demarchi, D. Guidotti, L. Pulina and A. Tacchella

      Supporting Standardization of Neural Networks Verification with VNN-LIB and CoCoNet Conference

      FoMLAS Workshop, Computer-Aided Verification

    • S. Demarchi

      Experimenting with Constraint Programming Techniques in Artificial Intelligence: Automated System Design and Verification of Neural Networks Thesis

      PhD Thesis

    • D. Guidotti, S. Demarchi, L. Pulina and A. Tacchella

      Evaluating Reachability Algorithms for Neural Networks on NeVer2 Preprint

      Submitted for peer review

  • 2022

    • D. Guidotti

      Verification and Repair of Machine Learning Models Thesis

      PhD Thesis

    • S. Demarchi and D. Guidotti

      Counter-Example Guided Abstract Refinement for Verification of Neural Networks Conference

      CPS Summer School Workshop, PhD workshop

    • S. Demarchi, D. Guidotti, A. Pitto and A. Tacchella

      Formal Verification of Neural Networks: a Case Study about Adaptive Cruise Control Conference

      International ECMS Conference on Modeling and Simulation

  • 2021

    • D. Guidotti

      Verification and Repair of Neural Networks Conference

      AAAI Conference on Artificial Intelligence

    • D. Guidotti, L. Pulina and A. Tacchella

      pyNeVer: a Framework for Learning and Verification of Neural Networks Conference

      Automated Technology for Verification and Analysis

    • D. Guidotti, G. Cicala, T. Gili and A. Tacchella

      Telling Faults From Cyber-Attacks In A Multi-Modal Logistic System With Complex Network Analysis Conference

      International ECMS Conference on Modeling and Simulation

    • D. Guidotti

      Safety Analysis of Deep Neural Networks Conference

      International Joint Conference on Artificial Intelligence (IJCAI)

  • 2020

    • D. Guidotti, F. Leofante, L. Pulina and A. Tacchella

      Verification of Neural Networks: Enhancing Scalability through Pruning Conference

      European Conference on Artificial Intelligence

  • 2012

    • L. Pulina and A. Tacchella

      Challenging SMT solvers to verifiy Neural Networks Article

      AI Communications

  • 2011

    • L. Pulina and A. Tacchella

      NeVer: a Tool for Artificial Neural Networks Verification Article

      Annals of Mathematics and Artificial Intelligence

    • L. Pulina and A. Tacchella

      Checking Safety of Neural Networks with SMT Solvers: A Comparative Evaluation Conference

      International Conference of AI*IA

  • 2010

    • L. Pulina and A. Tacchella

      An Abstraction-Refinement Approach to Verification of Artificial Neural Networks Conference

      Computer Aided Verification, 22nd International Conference

Team

...

Armando Tacchella

Full Professor | UniGE

...

Luca Pulina

Full Professor | UniSS

...

Dario Guidotti

Post-Doc Researcher | UniSS

...

Stefano Demarchi

Post-Doc Researcher | UniGE

...

Elena Botoeva

Lecturer | UniKent

...

Andrea Gimelli

Graduate Fellow | UniGE

...

Karim Pedemonte

Graduate Software Engineer | UniGE
Email
nevertools _at_ googlegroups.com
MIT Licensed | neuralverification.org 2024 | Hosted with ❤ by GitHub