Internet Explorer: Targeted Repr Learning on the Open Web

Stream: Machine Learning for Theorem Proving

[Junyan Xu]

we propose dynamically utilizing the Internet to quickly train a small-scale model that does extremely well on the task at hand. Our approach, called Internet Explorer, explores the web in a self-supervised manner to progressively find relevant examples that improve performance on a desired target dataset. It cycles between searching for images on the Internet with text queries, self-supervised training on downloaded images, determining which images were useful, and prioritizing what to search for next.

https://twitter.com/_akhaliq/status/1630413027227959296

Internet Explorer: Targeted Representation Learning on the Open Web abs: https://arxiv.org/abs/2302.14051 project page: https://internet-explorer-ssl.github.io/ https://twitter.com/_akhaliq/status/1630413027227959296/video/1

- AK (@_akhaliq)

Anyone to replace image search by literature search, vision model by language (or multimodal?) model, using mathlib as a starting point?