Artificial Intelligence
How Anthropic Internal Code Leaked via the Public npm Registry?
So, you’ve probably seen the news about the internal code exposure. It sounds like a sophisticated breach, but in reality, it was a classic…
Blog
Practical technical articles and clear explanations.
-
-
Artificial Intelligence
Understanding Pre-Training in Large Language Models
Pre-training is the phase where we teach a model how language works. Before a model can answer questions, write code, or chat with us,…
-
Artificial Intelligence
An Introduction to Vision Language Model
AI application now a days are not only generating texts , but also images, audio and videos. The similar approach of transformer architecture is…
-
Artificial Intelligence
Tokens and logits relation in LLM
A token is a piece of text the model understands. It may be: a word part of a word punctuation space Now each token…
-
Artificial Intelligence
Understanding Attention Heads in LLMs
The attention heads live inside the Transformer layer Each Transformer layer actually contains two main parts: 1 Multi-Head Self Attention2 Feed Forward…
-
Artificial Intelligence
How logits are formed in LLM?
The model scores (logits) come from the last neural network layer of the transformer. After many transformer layers, the model produces a vector that…
-
Artificial Intelligence
Softmax and how LLM is predicting the next word
How does an LLM pick the next word? A simple look at softmax, which turns raw scores into probabilities, with a small worked example.
-
Career
Elevating Your Professional Profile: From Task-Oriented to Value-Driven
In a competitive job market, your profile needs to be more than a chronological list of responsibilities. To truly stand out, you must demonstrate…
-
Artificial Intelligence
Epoch in model training
In the simplest terms, an Epoch represents one complete pass of your entire training dataset through the the model. Imagine you are training an…
-
Data Science
Chessboard distance aka Chebyshev distance
Chebyshev distance, or chessboard distance, explained simply: how a chess king moves, the formula, and where data science uses it.
