Yogeswari Lakshmiraman·Oct 5Machine Learning Series (Part 38) : From Class Scores to Probabilities with SoftmaxIn part 19 of our ML series, we discussed the different methodologies in training a multiclass classifier. Fundamentally, a multiclass…
InTowards AIbyDr Swarnendu AI·Jun 21Your Language Model Cannot Say Certain Sentences.There are next-word predictions your model is mathematically forbidden from making. Not unlikely. Forbidden, the way a piano with too few…A response icon2A response icon2
Yash Datta·May 15Streaming softmax: the recurrence that makes FlashAttention workA two-state recurrence (log-sum-exp with a running-max correction factor) turned 200K-token context windows from a hardware fantasy into a…
InTowards AIbyJumbong Junior Quantitatif analyst·May 10Learning Word Vectors for Sentiment Analysis: A Python ReproductionHow to build sentiment-aware word representations from IMDb reviews using semantic learning, star ratings, and linear SVM classification
Anuranjan Yadav·Mar 192.718: How This One Number Impacts What Your AI Says NextHave you ever wondered how an AI model, like ChatGPT, always seems to find the perfect next word in a sentence? Or how it knows to be more…
Uri Itai·Mar 16Move 37, Softmax, and Entropy: What a Single Go Move Revealed About Modern AI.In 2016, a single move in a game of Go shocked the world.A response icon1A response icon1
The Visionary Vectors Blog·Oct 15, 2025The Math I Need for TransformersBefore diving deep into LLMs, I wanted to get comfortable with the math behind them so the theory actually makes sense. So, I started…
Eshani Parulekar·Aug 30, 2025How Duolingo Keeps Me Learning French (and What We Can Learn From It)Last year, I challenged myself to start a 1,000-day Duolingo streak to finally take my French learning seriously. As I write this, I’m on…
