Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
ETH Robustness Analyzer for Deep Neural Networks
| Date | Stars |
|---|---|
| 2026-07-31 | 348 |
| 2026-08-05 | 348 |
| 2026-08-06 | 348 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
ERAN <img width="100" alt="portfolio_view" align="right" src="http://safeai.ethz.ch/img/sri-logo.svg"> ========  ETH Robustness Analyzer for Neural Networks (ERAN) is a state-of-the-art sound, precise, scalable, and extensible analyzer based on [abstract interpretation](https://en.wikipedia.org/wiki/Abstract_interpretation) for the complete and incomplete verification of MNIST, CIFAR10, and ACAS Xu based networks. ERAN produces state-of-the-art precision and performance for both complete and incomplete verification and can be tuned to provide best precision and scalability (see recommended configuration settings at the bottom). ERAN is developed at the [SRI Lab, Department of Computer Science, ETH Zurich](https://www.sri.inf.ethz.ch/) as part of the [Safe AI project](http://safeai.ethz.ch/). The goal of ERAN is to automatically verify safety properties of neural networks with feedforward, convolutional, and residual layers against input perturbations (e.g., L∞-norm attacks, geometric transformations, vector field deformations, etc). ERAN combines abstract domains with custom multi-neuron relaxations from PRIMA to support fully-connected, convolutional, and residual networks with ReLU, Sigmoid, Tanh, and Maxpool activations. ERAN is sound under floating point arithmetic with the exception of the (MI)LP solver used in RefineZono and RefinePoly. The employed abstract domains are specifically designed for the setting of neural networks and aim to balance scalability and precision. Specifically, ERAN supports the following analysis: * DeepZ [NIPS'18]: contains specialized abstract Zonotope transformers for handling ReLU, Sigmoid and Tanh activation functions. * DeepPoly [POPL'19]: based on a domain that combines floating point Polyhedra with Intervals. * GPUPoly [MLSys'2021]: leverages an efficient GPU implementation to scale DeepPoly to much larger networks. * RefineZono [ICLR'19]: combines DeepZ analysis with MILP and LP solvers for more precision. * RefinePoly/RefineGPUPoly [NeurIPS'19]: combines DeepPoly/GPUPoly analysis with (MI)LP refinement and PRIMA framework [arXiv'2021] to compute group-wise joint neuron abstractions for state-of-the-art precision and scalability. All analysis are implemented using the [ELINA](http://elina.ethz.ch/) library for numerical abstractions. More details can be found in the publications below. ERAN vs AI2 -------------------- Note that ERAN subsumes the first abstract interpretation based analyzer [AI2](https://www.sri.inf.ethz.ch/publications/gehr2018ai), so if you aim to compare, please use ERAN as a baseline. USER MANUAL -------------------- For a detailed desciption of the options provided and the implentation of ERAN, you can download the [user manual](https://files.sri.inf.ethz.ch/eran/docs/eran_manual.pdf). Requirements ------------ GNU C compiler, ELINA, Gurobi's Python interface, python3.6 or higher, tensorflow 1.11 or higher, numpy. Installation ------------ Clone the ERAN repository via git as follows: ``` git clone https://github.com/eth-sri/ERAN.git cd ERAN ``` The dependencies for ERAN can be installed step by step as follows (sudo rights might be required): Note that it might be required to use `sudo -E` to for the right environment variables to be set. Ensure that the following tools are available before using the install script: * cmake (>=3.17.1), * m4 (>=1.4.18) * autoconf, * libtool, * pdftex. On Ubuntu systems they can be installed using: ``` sudo apt-get install m4 sudo apt-get install build-essential sudo apt-get install autoconf sudo apt-get install libtool sudo apt-get install texlive-latex-base ``` Consult https://cmake.org/cmake/help/latest/command/install.html for the install of cmake or use: ``` wget https://github.com/Kitware/CMake/releases/download/v3.19.7/cmake-3.19.7-Linux-x86_64.sh sudo bash ./cmake-3.19.7-Linux-x86_64.sh sudo rm /usr/bin/cmake sudo ln -
Excerpt of 38,881 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:62f979f665ed9f94, llm:Project title/description: "ETH Robustness Analyzer for Deep Neural Networks" — a tool for analyzing robustness of DNNs (verification/robustness).
matched fp:62f979f665ed9f94, llm:Project title/description: "ETH Robustness Analyzer for Deep Neural Networks" — a tool for analyzing robustness of DNNs (verification/robustness).