0g
Neural network verifier
A system for proving an inference pass for pre-trained neural networks on private inputs.
40 stars
3 watching
2 forks
Language: Python
last commit: almost 2 years ago
Linked from 1 awesome list
Related projects:
| Repository | Description | Stars |
|---|---|---|
 zk-ml/tachikoma | Defines a standard for serializing neural network inference processes into a graph of operator computational traces. | 33 |
 worldcoin/proto-neural-zkp | A system to verify neural networks in zero-knowledge proofs using cryptographic techniques | 172 |
 lyronctk/zator | This project verifies the inference of a deep neural network using recursive SNARKs and leverages a folding scheme to reduce computation complexity. | 160 |
| zk-ml/uchikoma | A tool that translates neural network code into languages that can run on devices without floating-point operations. | 29 |
 vehicle-lang/vehicle | A toolkit for enforcing logical specifications on neural networks | 82 |
 priba/nmp_qc | An implementation of neural networks on graph structures for learning molecular properties | 340 |
 cedergrouphub/chgnet | A neural network potential for atomistic modeling | 258 |
 ahmedfgad/numpyann | An implementation of artificial neural networks using NumPy | 98 |
 zkonduit/ezkl | An engine for doing inference on deep learning models and computational graphs in a ZKML proof system | 970 |
| ahmedfgad/neuralgenetic | Trains artificial neural networks using the genetic algorithm | 241 |
 tamucrypto/zkcnn | An implementation of a GKR-based snark for CNN model inference, providing a zero-knowledge proof protocol for neural network computations. | 73 |
 cloudkj/layer | A utility for executing neural network computations through sequential command-line pipelines. | 558 |
 google-research/tf-slim | A lightweight library for defining and training neural networks in TensorFlow. | 373 |
 calculatedcontent/weightwatcher | A tool for analyzing Deep Neural Networks to predict accuracy and detect potential problems | 1,486 |
 kootenpv/neural_complete | A tool to aid in writing neural network code by suggesting completions based on learned patterns from similar source code | 1,151 |