Everything you care about in one place

Follow feeds: blogs, news, RSS and more. An effortless way to read and digest content of your choice.

Get Feeder


Off the convex path

Get the latest updates from Off the convex path directly as they happen.

Follow now 144 followers

Latest posts

Last updated 5 months ago

Ultra-Wide Deep Nets and Neural Tangent Kernel (NTK)

5 months ago

(Crossposted at CMU ML.) Traditional wisdom in machine learning holds that there...

Understanding implicit regularization in deep learning by analyzing trajectories of gradient descent

8 months ago

Sanjeev’s recent blog post suggested that the conventional view of optimization is...

Landscape Connectivity of Low Cost Solutions for Multilayer Nets

8 months ago

A big mystery about deep learning is how, in a highly nonconvex...

Is Optimization a Sufficient Language for Understanding Deep Learning?

9 months ago

In this Deep Learning era, machine learning usually boils down to defining...

Contrastive Unsupervised Learning of Semantic Representations: A Theoretical Framework

11 months ago

Semantic representations (aka semantic embeddings) of complicated data types (e.g. images, text...

The search for biologically plausible neural computation: A similarity-based approach

about 1 year ago

This is the second post in a series reviewing recent progress in...

Understanding optimization in deep learning by analyzing trajectories of gradient descent

over 1 year ago

Neural network optimization is fundamentally non-convex, and yet simple gradient-based algorithms seem...

Simple and efficient semantic embeddings for rare words, n-grams, and language features

over 1 year ago

Distributional methods for capturing meaning, such as word embeddings, often require observing...

When Recurrent Models Don't Need to be Recurrent

over 1 year ago

In the last few years, deep learning practitioners have proposed a litany...

Deep-learning-free Text and Sentence Embedding, Part 2

over 1 year ago

This post continues Sanjeev’s post and describes further attempts to construct elementary...

Deep-learning-free Text and Sentence Embedding, Part 1

over 1 year ago

Word embeddings (see my old post1 and post2) capture the idea that...

Limitations of Encoder-Decoder GAN architectures

almost 2 years ago

This is yet another post about Generative Adversarial Nets (GANs), and based...