I Built LLM Explained: 43 Interactive Pages That Take You From Tokens to AI Agents
I kept running into the same problem while learning about large language models. I'd read a blog post about attention, nod along, and two days later I couldn't tell you why we divide by √d_k. I'd see "GQA", "MLA", "MoE", "LoRA", "HNSW" or "context rot" in a paper or a tweet, open five tabs, and close them all with the uneasy feeling that I sort of got it. And the field keeps moving, so there's…
Amanda Nair created LLM Explained, a free interactive course that takes learners from the fundamentals of language models to advanced AI agents. The resource comprises 42 modules across seven parts, totaling 43 interactive pages. Each concept is presented visually through interactive animations, followed by mathematical derivations for those who wish to delve deeper.
The course emphasizes a "picture first, symbol second" approach, with every equation accompanied by an animation that mirrors the colors used in the diagrams. This ensures that learners can easily correlate visual representations with their corresponding equations. Modules are designed with bite-sized information, moving forward as you scroll without any automated scrolling.
Interactive elements include at least one customizable widget per module, allowing users to manipulate parameters and observe the effects in real-time. These widgets provide hands-on experience, enabling learners to experiment with different settings and observe the outcomes immediately. Additionally, every module concludes with a "What this picture leaves out" section, highlighting the limitations of simplified diagrams to prevent misconceptions.
All numerical values displayed on the site are computed on the user's browser, ensuring transparency and accuracy.
To navigate the course, users can follow a curated curriculum map or select modules based on their depth preference: Intuition (pictures and plain language), Mechanics (tensor shapes and step-by-step operations), or Maths (full derivations). These three levels allow learners to progress at their own pace, revisiting content as needed. The course also includes a feature that tracks the number of parameters against established benchmarks, facilitating comparisons and a deeper understanding of model complexity.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
