Skip to content
#

twitter-algorithm

Here are 17 public repositories matching this topic...

A transparent, tunable recommendation engine reverse-engineering the X (Twitter) algorithm. Features multi-objective candidate ranking, semantic cosine embeddings, explainable scoring breakdowns, real-time parameter tuning, and a multi-agent social simulation environment built with FastAPI and Next.js.

  • Updated Aug 20, 2026
  • TypeScript

An unofficial, full source-read analysis of X's open-sourced "For You" recommendation algorithm - ML ranker, retrieval, Rust pipeline, and Grok-LLM moderation explained for both engineers and everyday users.

  • Updated Jul 20, 2026
  • Rust

Growth 2.0: an autonomous evolutionary loop that hacks Twitter's post-December-2025 ranking algorithm. Karpathy autoresearch × EvoHarness tree search × offline Twitter simulator × Grok opponent study. Encodes lessons from Oyster Labs' Growth 1.0 bootstrap — a hunger game of content strategies inside a simulation arena.

  • Updated Apr 18, 2026
  • Python

Add this topic to your repo

To associate your repository with the twitter-algorithm topic, visit your repo's landing page and select "manage topics."

Learn more