FILM
Gradient attacks
Recovering private text in federated learning of language models by identifying and reconstructing sensitive client data from model gradients
Official repo for the paper: Recovering Private Text in Federated Learning of Language Models (in NeurIPS 2022)
57 stars
4 watching
7 forks
Language: Python
last commit: over 1 year ago Related projects:
| Repository | Description | Stars |
|---|---|---|
 dcalab-unipv/turning-privacy-preserving-mechanisms-against-federated-learning | This project presents an attack on federated learning systems to compromise their privacy-preserving mechanisms. | 8 |
 jeremy313/fl-wbc | A defense mechanism against model poisoning attacks in federated learning | 37 |
 lhfowl/robbing_the_fed | This implementation allows an attacker to directly obtain user data from federated learning gradient updates by modifying the shared model architecture. | 23 |
 ftramer/steal-ml | An implementation of extraction attacks against Machine Learning models offered by Cloud-based services | 344 |
 hfzhang31/a3fl | A framework for attacking federated learning systems with adaptive backdoor attacks | 22 |
 git-disl/lockdown | A backdoor defense system against attacks in federated learning algorithms used for machine learning model training on distributed datasets. | 14 |
 jonasgeiping/breaching | A PyTorch framework for analyzing vulnerabilities in federated learning models and predicting data breaches | 269 |
 yunqing-me/attackvlm | An adversarial attack framework on large vision-language models | 161 |
| jeremy313/soteria | An implementation of a defense against model inversion attacks in federated learning | 55 |
 ibm/reprogrammble-fl | Improves utility-privacy tradeoff in federated learning by reprogramming models to balance data utility and user privacy. | 5 |
 ymjs-irfan/dp-fedsam | This repository provides an implementation of a differentially private federated learning algorithm designed to improve the robustness and performance of federated machine learning systems. | 41 |
 facebookresearch/spiritlm | This repository provides an end-to-end language model capable of generating coherent text based on both spoken and written inputs. | 777 |
 ai-secure/dba | A tool for demonstrating and analyzing attacks on federated learning systems by introducing backdoors into distributed machine learning models. | 176 |
 sap-samples/machine-learning-diff-private-federated-learning | Simulates a federated learning setting to preserve individual data privacy | 360 |
 umbc-sanjaylab/fedpseudo_kdd23 | This repository provides an implementation of federated survival analysis using a deep learning framework. | 0 |