So far, we've learned why embeddings exist, how vectors, dimensions, and similarity work, how nn.Embedding stores vectors and retrieves them using token IDs, how padding and attention masks handle var
In Module 1, we learned why embeddings exist. In Module 2, we understood what vectors are, including their dimensions, magnitude, dot products, and cosine similarity. Now, let's connect everything to
In Module 1, we learned that an embedding layer converts a token ID into a vector. For example: cat = [0.90, 0.10] dog = [0.85, 0.15] We said these two vectors look similar, but we haven't explained w
In our Inside an LLM Tokenizer , we understand Unicode → UTF-8 → Tokenization → Token IDs Now, the next question is: How does an LLM turn those meaningless integer IDs into numerical representations t
Go to https://platform.openai.com/tokenizer Put 🏡 in the text box. And see the output. Why one single emoji is giving us two tokenIDs? For that, we need to understand three different things: Uni
The strange pause Okay, so lets say, you open a chat app and you paste a long document. Then, you ask, "Hey, can you summarize this?" You hit enter. Nothing happens. One second. Two seconds. Maybe thr
Introduction In Part 1, we built a RAG agent that can think and fix its own search queries. In Part 2, we gave it structured memory using hierarchical indexing and hybrid storage. We now have a capable researcher. It can handle vague questions. It ca...
Introduction In Part 1, we gave our RAG system a brain. We built an agent that can analyze queries, ask for clarification, and fix its own mistakes using LangGraph. But even the smartest agent is useless if it’s reading from a messy library. Today, w...
Introduction Retrieval-Augmented Generation (RAG) has become the "Hello World" of modern GenAI development. It is the standard way to ground Large Language Models (LLMs) in your own data. But let’s be honest—most RAG pipelines are "dumb." They are fr...
