∑

Category: Machine Learning

Showing 28 article(s)

AI Model Distillation: Principles, Methods, Advantages, and Limitations
Artificial IntelligenceMachine Learning

AI Model Distillation: Principles, Methods, Advantages, and Limitations

Knowledge distillation (or model distillation) is an AI compression technique where a smaller, faster model called the 'student model' (Student Model) is trained to reproduce the behavior of a more complex model called the 'teacher model' (Teacher Model).

6 min read
Why is the AI revolution so slow? (It’s not)
Artificial IntelligenceMachine Learning

Why is the AI revolution so slow? (It’s not)

Is the AI revolution really slow? History shows otherwise. In fact, it is the fastest technological change in history. Let's dive into this topic.

7 min read
How Does LLMs Store Knowledge? A Deep Dive Into Feature Superposition
Artificial IntelligenceMachine Learning

How Does LLMs Store Knowledge? A Deep Dive Into Feature Superposition

Ask ChatGPT about quantum physics, medieval history, or cooking, and it delivers precise answers, even offline. How does it know so much? The secret is feature superposition, a mechanism allowing AI to compress vast knowledge into a finite space. This deep dive explores how AI stores knowledge using this fascinating property. The Foundation: Disentangling Features […]

9 min read
How RLHF works for LLMs : A Deep Dive
Artificial IntelligenceMachine Learning

How RLHF works for LLMs : A Deep Dive

A deep dive into Reinforcement Learning with Human Feedback (RLHF) for LLMs. The key to make LLMs practical and aligned with human values.

9 min read
Attention Mechanism in LLM Explained : A Deep Dive
Artificial IntelligenceMachine Learning

Attention Mechanism in LLM Explained : A Deep Dive

Deep dive into the attention mechanism powering Transformer LLMs: understand the Query-Key-Value idea and discover efficient implementations.

11 min read
Tokenization in LLMs: Why Not Use Words?
Artificial IntelligenceMachine Learning

Tokenization in LLMs: Why Not Use Words?

Why do LLM use tokens and not words? The answer : we tried and it wasn't good. Explore with me how Tokenization evolved over time.

5 min read
Defining AGI : Why OpenAI’s o3 Isn’t Enough to achieve Artificial General Intelligence
Artificial IntelligenceMachine Learning

Defining AGI : Why OpenAI’s o3 Isn’t Enough to achieve Artificial General Intelligence

How close are we to AGI? Why OpenAI’s O3 sparks debate and discover the challenges in defining and achieving Artificial General Intelligence.

7 min read
The History of ChatGPT and Large Language Models : A timeline
Artificial IntelligenceMachine Learning

The History of ChatGPT and Large Language Models : A timeline

Discover ChatGPT’s and LLM history from 1950 to today in 8 key milestones. From AlexNet and Word2Vec to GPT and more!

5 min read
Why Use Non linearity in Artificial Neural Networks? An Intuitive Explanation
Artificial IntelligenceMachine Learning

Why Use Non linearity in Artificial Neural Networks? An Intuitive Explanation

Discover why non-linearity (ReLU, sigmoid, tanh) is crucial for separating classes, with simple, visualized examples.

3 min read