Code associated with ICML (2024). "Defense against Backdoor Attack on Pre-trained Language Models via Head Pruning and Attention Normalization"
☆10Feb 22, 2026Updated 2 weeks ago
Alternatives and similar repositories for PURE
Users that are interested in PURE are comparing it to the libraries listed below
Sorting:
- ☆15Dec 12, 2023Updated 2 years ago
- ☆18Aug 15, 2022Updated 3 years ago
- TextGuard: Provable Defense against Backdoor Attacks on Text Classification☆14Nov 7, 2023Updated 2 years ago
- [NDSS 2025] "CLIBE: Detecting Dynamic Backdoors in Transformer-based NLP Models"☆25Aug 20, 2025Updated 6 months ago
- ☆26Dec 1, 2022Updated 3 years ago
- ☆28Aug 21, 2023Updated 2 years ago
- An open-source toolkit for textual backdoor attack and defense (NeurIPS 2022 D&B, Spotlight)☆200Apr 10, 2023Updated 2 years ago
- ☆13Oct 4, 2024Updated last year
- ☆12Aug 6, 2024Updated last year
- Reconstructive Neuron Pruning for Backdoor Defense (ICML 2023)☆39Dec 24, 2023Updated 2 years ago
- [ICML 2023] Protecting Language Generation Models via Invisible Watermarking☆13Sep 8, 2023Updated 2 years ago
- Adversarial Attack for Pre-trained Code Models☆10Jul 19, 2022Updated 3 years ago
- [NeurIPS'24] Official implement of "PrivCirNet: Efficient Private Inference via Block Circulant Transformation"☆15Feb 26, 2026Updated last week
- ESEC/FSE'21: Prediction-Preserving Program Simplification☆10Oct 4, 2022Updated 3 years ago
- now-defunct fork of three20 -- please see facebook/three20 for most/all purposes☆17Aug 20, 2010Updated 15 years ago
- ☆14Feb 26, 2025Updated last year
- Implementation for the protocols described in https://eprint.iacr.org/2023/1700☆14Jan 9, 2025Updated last year
- Code for the paper: Fast and Private Inference of Deep Neural Networks by Co-designing Activation Functions☆11Mar 13, 2024Updated last year
- Your finetuned model's back to its original safety standards faster than you can say "SafetyLock"!☆11Oct 16, 2024Updated last year
- Two-party Privacy-preserving Neural Network Training using Split Learning and Homomorphic Encryption (CKKS Scheme)☆11Sep 23, 2025Updated 5 months ago
- The dataset, source code and the results of our ESEC/FSE 2023 paper "An Extensive Study on Adversarial Attack against Pre-trained Models …☆15Feb 19, 2024Updated 2 years ago
- This repository contains the code, the dataset and the experimental results related to the paper "Vulnerabilities in AI Code Generators: …☆12Aug 5, 2024Updated last year
- In this project, we used 3 different metrics (Information Gain, Mutual Information, Chi Squared) to find important words and then we used…☆13Aug 7, 2018Updated 7 years ago
- [ICLR 2025] On Evluating the Durability of Safegurads for Open-Weight LLMs☆13Jun 20, 2025Updated 8 months ago
- All-day Semantic Segmentation & All-day CityScapes dataset☆13Jul 26, 2025Updated 7 months ago
- Code release for MPCViT accepted by ICCV 2023☆16Jan 6, 2025Updated last year
- ☆14May 23, 2023Updated 2 years ago
- Fork from http://github.com/react-component/calendar.git☆11Feb 24, 2016Updated 10 years ago
- The implementation of the IEEE S&P 2024 paper MM-BD: Post-Training Detection of Backdoor Attacks with Arbitrary Backdoor Pattern Types Us…☆16May 12, 2024Updated last year
- Single Threaded CPU Scheduling Simulator☆11Dec 21, 2017Updated 8 years ago
- [ICLR 2025] REFINE: Inversion-Free Backdoor Defense via Model Reprogramming☆13Feb 13, 2025Updated last year
- ☆11Feb 21, 2022Updated 4 years ago
- Backdooring Neural Code Search☆14Sep 8, 2023Updated 2 years ago
- A repository containing the code for the paper "Incorporating Domain Knowledge into Medical NLI using Knowledge Graphs" EMNLP 2019☆13Nov 2, 2019Updated 6 years ago
- llama-70b-chat-4-shards implementation☆13Jun 4, 2024Updated last year
- ☆19Nov 27, 2024Updated last year
- a pure front-end upyun form upload service, supports both native js and angular.js☆30Jan 4, 2015Updated 11 years ago
- Discord爬蟲，把Discord頻道所有內容全部抓下來儲存☆12Jan 31, 2021Updated 5 years ago
- An evaluation framework for mitigating DNN backdoor attacks using data augmentations☆11Dec 10, 2020Updated 5 years ago