# TP4

**Repository Path**: only_zy/TP4

## Basic Information

- **Project Name**: TP4
- **Description**: https://arxiv.org/abs/2011.14522  官方代码库 仅搬运
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-12-16
- **Last Updated**: 2021-12-16

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

Empirical Experiments in "Feature Learning in Infinite-width Neural Networks"
========

This repo contains code to replicate our experiments (Word2Vec, MAML) in our paper

[**Feature Learning in Infinite-Width Neural Networks**](https://arxiv.org/abs/2011.14522)<br>
*Greg Yang, Edward Hu*

In short, the code here will allow you to train feature learning infinite-width neural networks on Word2Vec and on Omniglot (via MAML).

Our results on Word2Vec:

<img src="word2vec.png" alt="Word2Vec Results" width="325"/>


Our Results on MAML:

<img src="MAML.png" alt="MAML Results" width="200"/>


Please see the README in individual folders for more details.

This is the 4th paper in the Tensor Programs series ([[0]](http://arxiv.org/abs/1902.04760)[[1]](http://arxiv.org/abs/1910.12478)[[2]](http://arxiv.org/abs/2006.14548)[[3]](http://arxiv.org/abs/2009.10685)). Also see here for code in previous papers for calculating the [GP](https://github.com/thegregyang/GP4A) and [NTK](https://github.com/thegregyang/NTK4A) limits of wide neural networks.