Depths of Arxiv (@DepthsofArxiv) header
Depths of Arxiv (@DepthsofArxiv) profile picture

The quirkiest papers/statements/LaTeX snippets on Arxiv. DMs open for submission.

Joined July 2022
Photo by Depths of Arxiv (@DepthsofArxiv)Photo by Depths of Arxiv (@DepthsofArxiv)Photo by Depths of Arxiv (@DepthsofArxiv)Photo by Depths of Arxiv (@DepthsofArxiv)Photo by Depths of Arxiv (@DepthsofArxiv)Photo by Depths of Arxiv (@DepthsofArxiv)Photo by Depths of Arxiv (@DepthsofArxiv)Photo by Depths of Arxiv (@DepthsofArxiv)Photo by Depths of Arxiv (@DepthsofArxiv)
We are one of the best AI labs in the world but we will keep the rest of the world behind by a year by simply not releasing a preprint.
1
7
Title: Discovering faster matrix multiplication algorithms with reinforcement learning Source: nature.com/articles/s41586-0…
No models were harmed in the making of this preprint. Source: ofir.io/self-ask.pdf Authors: @OfirPress muruzhang @sewon__min ludwigschmit noahsmith @ml_perception
9
The todo is so heavy, it makes a sound. Source: arxiv.org/abs/1211.3711 Author: Alex Graves
15
Cultivating your inner reviewer #2
1
1
4
Title: Attend, Infer, Repeat: Fast Scene Understanding with Generative Models Source:arxiv.org/e-print/1603.08575 Authors: @theophaneweber @yuvaltassa @koraykv
Why waste time say lot word when few word do trick. Yoshua Bengio (YB) replying to a co-author (perhaps @goodfellow_ian ?) of Generative Adversarial Nets.
2
2
5
59