Volume XIV • Technical Archive

Articles

A curated collection of technical explorations, architectural insights, and research findings from my exploration and work with new technologies.

search
The AI Compass: Gradients & The Mountain of Loss
AI/ML
January 2026 2 min read

The AI Compass: Gradients & The Mountain of Loss

Understanding how gradients guide AI models down the mountain of loss toward better predictions.

Machine Learning Gradients AI Tutorial
AI/ML
William Golovlev arrow_forward
The Radio Knobs of AI: Weights, Biases, and the Learning Loop
AI/ML
January 2026 3 min read

The Radio Knobs of AI: Weights, Biases, and the Learning Loop

Understanding how AI models actually learn through weights, biases, gradients, and the training loop.

Machine Learning Weights Biases
AI/ML
William Golovlev arrow_forward
The Atoms of AI: Understanding Tensors
AI/ML
January 2026 2 min read

The Atoms of AI: Understanding Tensors

Breaking down what tensors actually are and why they're the foundation of machine learning.

Machine Learning Tensors PyTorch
AI/ML
William Golovlev arrow_forward
AI Loss Functions: How a Model Knows It Is Wrong
AI/ML
January 2026 2 min read

AI Loss Functions: How a Model Knows It Is Wrong

Understanding how AI models measure their mistakes and learn from them.

Machine Learning Loss Functions AI Tutorial
AI/ML
William Golovlev arrow_forward
Why Your AI is Probably Cheating (and How to Stop It)
AI/ML
January 2026 1 min read

Why Your AI is Probably Cheating (and How to Stop It)

Diving into One-Hot encoding, smart data splitting, and the trap of the single metric.

Machine Learning AI Ethics Data Science
AI/ML
William Golovlev arrow_forward
How do your favorite LLMs like ChatGPT work, and where is the future for LLMs heading?
AI/ML
January 2026 1 min read

How do your favorite LLMs like ChatGPT work, and where is the future for LLMs heading?

How neural networks use embeddings to conquer NLP and tabular data.

NLP LLM ChatGPT
AI/ML
William Golovlev arrow_forward
arrow_back Previous
2 of 4
Next arrow_forward