Twitter describes how to design local and computationally efficient provably powerful graph neural networks.

Twitter shares a technical analysis of its assessment for potential bias in its image cropping algorithm as part of its efforts to be more transparent around how it uses machine learning to improve pe

Twitter engineers found and fixed a Linux kernel bug in memory shrinker which caused OOM for us.

This blog post is based on Michael Bronstein’s keynote talk at ICLR 2021 and the paper M. M. Bronstein et al., Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges (2021)

Twitter annotated its data to a defined taxonomy by leveraging machine learning in its data platform by fusing Elasticsearch with neural networks.

Learn about Twitter’s adoption of RocksDB within Manhattan, its internal distributed key-value store.

In this post, we describe a graph neural network architecture (SIGN) that is of simple implementation and that works on very large graphs.

Introducing the dict-minimize package

Learn how Shardlib, a new sharding library at Twitter simplifies the management of sharded ads services and enables dynamic resharding of these services without redeploying their clients.

Twitter built a data product to help with coronavirus research by opening our APIs to academic researchers, to access the public conversation around COVID-19.