Digests » 125
this week's favorite
Transformers are Ruining Convolutions. This paper, under review at ICLR, shows that given enough data, a standard Transformer can outperform Convolutional Neural Networks in image recognition tasks, which are classically tasks where CNNs excel. In this Video, I explain the architecture of the Vision Transformer (ViT), the reason why it works better and rant about why double-bline peer review is broken.
The feature store is a data warehouse of features for machine learning (ML). Architecturally, it differs from the traditional data warehouse in that it is a dual-database, with one database (row-oriented) serving features at low latency to online applications and the other database (column-oriented) storing large volumes of features, used by Data Scientists to create train/test datasets and by batch applications doing offline model scoring.
A team of scientists at LMU Munich have developed Pattern-Exploiting Training (PET), a deep-learning training technique for natural language processing (NLP) models. Using PET, the team trained a Transformer NLP model with 223M parameters that out-performed the 175B-parameter GPT-3 by over 3 percentage points on the SuperGLUE benchmark.
An overview of concepts from machine learning and data science interviews presented as a series of tutorials along with practice questions at the end of each section.
Artificial intelligence in its current state is unfair, easily susceptible to attacks and notoriously difficult to control. Often, AI systems and predictions amplify existing systematic biases even when the data is balanced. Nevertheless, more and more concerning the uses of AI technology are appearing in the wild. This list aims to track all of them. We hope that Awful AI can be a platform to spur discussion for the development of possible preventive technology (to fight back!).