Papers With Code Adds State-of-the-Art Features

Papers With Code is a unique and useful resource that presents trending ML research along with the code to implement it. The site was created by Atlas ML CEO Robert Stojnic, aka “rstoj” on Reddit’s machine learning board. The latest version of Papers With Code has added 950+ unique machine learning tasks, 500+ State-of-the-Art result leaderboards and 8500+ papers with code. Papers With Code also enables users to match a machine learning paper on arXiv to its code on GitHub, which can help with reviewing content from different perspectives.

Data Source

Most of the Papers With Code data comes from the site’s own annotation team. The site has coverage on over 60,000+ papers, including most machine learning tasks. Papers With Code has also manually categorized tasks and Data Source indicators for more than 1,600 arXiv paper abstracts from the last three months of 2018. The Papers to Code team has also leveraged data from the following open source and free license projects:

• NLP progress
• EFF AI metrics
• SQuAD
• RedditSota

Related OFFICIAL Open-Source data downloading and code

Data Downloading
(All the data is licensed under the CC BY-SA licence)

• Papers with abstracts
• Links between papers and code
• Evaluation tables

SOTA Extractor Pipeline

• Code to Scrape and Import Data

Everything on the Papers to Code site is editable and versioned. Stojnic outlined his hopes for the upgraded platform in a recent Reddit post: “We’ve found the tasks and state-of-the-art data really informative to discover and compare research – and even found some research gems that we didn’t know about before. Feel free to join us in annotating and discussing papers!”

---

Author: Robert Tian | Editor: Michael Sarazen