Home

Interactive research curriculum · 2026

LLM & VLM Research Learning Lab

88 planned sessions across 14 survey-guided research tracks: Transformers, distributed training, fine-tuning, efficient inference, VLMs, video understanding, retrieval and diffusion. Existing 46 pages remain accessible; four new interactive chapters are in development. Planned sessions are clearly labeled.

Explore 14 research tracks →
📚 All sessionsBrowse the complete curriculum 🧠 Transformer anatomyInteractive architecture exploration 👁️ Vision-language modelsEncoder · projector · decoder 🎞️ Video retrievalSearch video using embeddings

Learning Muse · Previous Articles

Preface:

Hi, I’m Mehdi, and I use this blog to document my learning journey. I am a research scientist at Meta | Exploring computer vision, computational photography, and Gen-AI 🚀. I’ve discovered that I grasp concepts more effectively when I understand a simple code example associated with them. Therefore, in this blog, I explore new concepts through coding!

I try to use simple scenarios and of course, I never use ChatGPT .





  • Click on each icon below for categorical topics.