🐢 Open-Source Evaluation & Testing for ML & LLM systems

Corruption and Perturbation Robustness (ICLR 2019)

A Harder ImageNet Test Set (CVPR 2021)

Deep Anomaly Detection with Outlier Exposure (ICLR 2019)

Deliver safe & effective language models

PromptInject is a framework that assembles prompts in a modular fashion to provide a quantitative analysis of the robustness of LLMs to adversarial prompt attacks. 🏆 Best Paper Awards @ NeurIPS ML Safety Workshop 2022

Self-Supervised Learning for OOD Detection (NeurIPS 2019)

ImageNet-R(endition) and DeepAugment (ICCV 2021)

Aligning AI With Shared Human Values (ICLR 2021)

Repo for "Benchmarking Robustness of 3D Point Cloud Recognition against Common Corruptions" https://arxiv.org/abs/2201.12296

📚 A curated list of papers & technical articles on AI Quality & Safety

The Combined Anomalous Object Segmentation (CAOS) Benchmark

Pre-Training Buys Better Robustness and Uncertainty Estimates (ICML 2019)

[ICML 2019] ME-Net: Towards Effective Adversarial Robustness with Matrix Estimation

Jiminy Cricket Environment (NeurIPS 2021)

Predicting Out-of-Distribution Error with the Projection Norm

Evaluation & testing framework for computer vision models

Code for the attack multiplicative filter attack MUFIA, from the paper "Frequency-based vulnerability analysis of deep learning models against image corruptions".

This repository contains the project for the Advanced AI course @CentraleSupélec