<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom" ><generator uri="https://jekyllrb.com/" version="3.10.0">Jekyll</generator><link href="https://facebookresearch.github.io/maws/feed.xml" rel="self" type="application/atom+xml" /><link href="https://facebookresearch.github.io/maws/" rel="alternate" type="text/html" /><updated>2025-04-13T02:42:16+00:00</updated><id>https://facebookresearch.github.io/maws/feed.xml</id><title type="html">The effectiveness of MAE pre-pretraining for billion-scale pretraining</title><subtitle>Project website for the paper &quot;The effectiveness of MAE pre-pretraining for  billion-scale pretraining&quot;. Contains links to an interactive Colab notebook with a multilingual zero shot demo, github link to the released MAE (Masked Autoencoder) and MAWS (MAE-&gt;WSP) models, and link to the paper on arxiv. </subtitle></feed>