Skip to content
  • Categories
  • CSPaper Review
  • Recent
  • Tags
  • Popular
  • World
  • Paper Copilot
  • OpenReview.net
  • Deadlines
  • CSRanking
Skins
  • Light
  • Cerulean
  • Cosmo
  • Flatly
  • Journal
  • Litera
  • Lumen
  • Lux
  • Materia
  • Minty
  • Morph
  • Pulse
  • Sandstone
  • Simplex
  • Sketchy
  • Spacelab
  • United
  • Yeti
  • Zephyr
  • Dark
  • Cyborg
  • Darkly
  • Quartz
  • Slate
  • Solar
  • Superhero
  • Vapor

  • Default (No Skin)
  • No Skin
Collapse
CSPaper

CSPaper: peer review sidekick

  1. Home
  2. Uncategorized
  3. Paper With Code Redirects to GitHub — What It Means for CS Researchers

Paper With Code Redirects to GitHub — What It Means for CS Researchers

Scheduled Pinned Locked Moved Uncategorized
paper with codegithubdownreproducibilityblogdiscoverabilitytransparencyresearchcomputer science
1 Posts 1 Posters 188 Views
  • Oldest to Newest
  • Newest to Oldest
  • Most Votes
Reply
  • Reply as topic
Log in to reply
This topic has been deleted. Only users with topic management privileges can see it.
  • cqsyfC Offline
    cqsyfC Offline
    cqsyf
    Super Users
    wrote last edited by
    #1

    In a surprising turn of events, the widely used platform PapersWithCode has effectively disappeared, now redirecting users to its GitHub organization page. Many researchers noticed this shift without any prior announcement, sparking confusion and concern in the community.

    teaser-paperwcode-redirects-github.jpg

    The sudden redirection was first noticed on July 26, 2025. Even bookmarks to the main domain now open GitHub directly, and all search functionality, leaderboard tracking, and paper-code linking features from the web interface are gone.

    “This morning I wanted to look up the code for a paper, but to my shock, PaperWithCode redirected to GitHub. I thought I mistyped the URL.”
    — User on WeChat blog, July 26, 2025


    🧠 Why It Matters

    1. Loss of a Central Hub for Reproducible ML Research

    Papers with Code wasn’t just a site for convenience — it was a crucial infrastructure for the machine learning and computer vision community. By aggregating:

    • research papers from arXiv,
    • official code repositories,
    • evaluation metrics and benchmarks,
    • leaderboard tracking (e.g., SOTA on ImageNet, GLUE, etc.),

    it offered a powerful one-stop shop for transparent, reproducible, and benchmarked research.

    Its loss is more than symbolic — it’s a hit to the culture of open research that has flourished over the past decade.


    2. The “Code” in “PaperWithCode” Was the Priority

    “The most important thing in computer science is code, not paper. Paper without code is just... well, you know.”
    — User commentary on GitHub issues

    This reflects a growing sentiment in the CS community — that reproducibility and implementation are as important as theoretical novelty. PWCode played a pivotal role in reinforcing this principle by nudging authors toward publishing their code alongside papers.

    paper-with-code-rediretcs-github-issue.jpeg


    3. Redirection to GitHub ≠ Replacement

    While GitHub hosts the back-end code and datasets, it lacks:

    • search capabilities for papers by metric or task,
    • benchmark SOTA tracking,
    • automatic arXiv-paper-to-code linking,
    • and user-friendly navigation for non-developer researchers.

    The redirection strips away all the curated glue that made the platform a discovery engine for ML developments.


    ❓ Unanswered Questions

    • Was this shut down planned?
      There was no official notice or blog post from the PapersWithCode team.

    • Is it permanent or temporary?
      With no service status or roadmap shared publicly, it remains unclear whether this is a technical issue or a deliberate sunset of the platform.

    • Why are there so few GitHub issues?
      Some users noted the surprisingly low number of support tickets filed on GitHub. This could indicate a drop in engagement, or simply that many users are unaware of where or how to respond.


    💭 Final Thoughts

    The sudden disappearance of PapersWithCode is a blow to the CS research community — especially to those who valued discoverability, transparency, and reproducibility.

    If this shift becomes permanent, we may see a resurgence of siloed research and increased difficulty for newcomers trying to replicate or build on state-of-the-art work.

    Let’s hope the team behind PWCode, or the broader open research community, steps up with a new initiative or public roadmap soon.


    Have thoughts or suggestions for alternatives to PaperWithCode? Share them in the comments.

    1 Reply Last reply
    0
    Reply
    • Reply as topic
    Log in to reply
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes


    • Login

    • Don't have an account? Register

    • Login or register to search.
    © 2025 CSPaper.org Sidekick of Peer Reviews
    Debating the highs and lows of peer review in computer science.
    • First post
      Last post
    0
    • Categories
    • CSPaper Review
    • Recent
    • Tags
    • Popular
    • World
    • Paper Copilot
    • OpenReview.net
    • Deadlines
    • CSRanking