REDDIT · REDDIT// 5h agoTUTORIAL

Anonymous Github surfaces for code review

A Reddit discussion on anonymizing code for AI/ML conference submissions quickly converged on Anonymous Github, a 4open.science tool that mirrors GitHub repositories while hiding owner, organization, repo, file, and content identifiers. The practical answer: use a dedicated anonymization service or anonymized supplementary archive, then restore official attribution after acceptance according to conference rules.

// ANALYSIS

This is less a product launch than a recurring pain point in double-blind ML review: code is now expected, but GitHub identity leaks remain easy to miss.

–Anonymous Github is useful because it preserves a browsable repo interface instead of forcing reviewers through a zip file
–Authors still need to scrub hardcoded usernames, lab names, paths, affiliations, wandb projects, dataset URLs, and commit metadata
–Conference policy matters more than community convention; some venues prefer supplementary files, some tolerate anonymous mirrors, and some warn that identifying code can trigger rejection
–The safest workflow is to freeze an anonymized review artifact at submission time, then move or relink to the official repository after acceptance

// TAGS

anonymous-githubopen-sourcedevtoolresearchmlops

DISCOVERED

5h ago

2026-04-22

PUBLISHED

6h ago

2026-04-22

RELEVANCE

5/ 10

AUTHOR

Terrible-Chicken-426