Create an AI capable of solving reasoning tasks it has never seen before

Official implementation of NeurIPS'23 paper "Truncated Affinity Maximization: One-class Homophily Modeling for Graph Anomaly Detection"

Official implementation for NeurIPS'24 paper "Generative Semi-supervised Graph Anomaly Detection"

Data Effiecient Image Transformers Implemented on a smaller Scale, using the CIFAR100 dataset and downsampled Imagenet ImageNet32

A collection of resources regarding the interplay between differential equations, deep learning, dynamical systems, control and numerical methods.

Experiments from the paper "On Second Order Behaviour in Augmented Neural ODEs"

Effortless plugin and play Optimizer to cut model training costs by 50%. New optimizer that is 2x faster than Adam on LLMs.

ICIP 2023: Pancreas Segmentation

Weight-Averaged Sharpness-Aware Minimization (NeurIPS 2022)

Official Pytorch implementation of paper "Zero Stability Well Predicts Performance of Convolutional Neural Networks"

Code for the paper "Robustness Requires Different Structures Than Residual Connection"

Understanding Adversarial Robustness of Vision Transformers via Cauchy Problem - ECML 2022 & Software Impacts (2023)

Official repo for consistency models.

This is a list of peer-reviewed representative papers on deep learning dynamics (optimization dynamics of neural networks). The success of deep learning attributes to both network architecture and …

Low Precision Arithmetic Simulation in PyTorch

Visualzation methods that help developers to realize the deep network (Explainable AI)

大学研究生院雨课堂的脚本仓库，该仓库下的脚本经过小改动后也适用于其他院校的雨课堂网课作业和视频。

Create animations for the optimization trajectory of neural nets

Let's train vision transformers (ViT) for cifar 10!

Train to 94% on CIFAR-10 in <6.3 seconds on a single A100. Or ~95.79% in ~110 seconds (or less!)

Official Open Source code for "Masked Autoencoders As Spatiotemporal Learners"

Deformable ConvNets V2 (DCNv2) in PyTorch

A collection of resources and papers on Diffusion Models

The Modified Differential Multiplier Method (MDMM) for PyTorch

[NeurIPS 2022] A novel 1-Lipschitz network that can be efficiently trained to achieve certified L-infinity robustness for free!

functorch is JAX-like composable function transforms for PyTorch.

LeViT a Vision Transformer in ConvNet's Clothing for Faster Inference

A PyTorch library entirely dedicated to neural differential equations, implicit models and related numerical methods

Port-Hamiltonian Approach to Neural Network Training