In an LLM, as in evolution, complexity emerges from simplicity. The computations that a transformer performs are relatively simple, involving the embedding of feature vectors, their weighting with self-attention, and the distribution of computation across heads and layers.

— Christopher Summerfield, These Strange New Minds, 163-164.

A short glossary with cites, which admittedly does not help a whole lot. I am still waiting for the light to go on:

Transformer
“A transformer model is a neural network that learns context and thus meaning by tracking relationships in sequential data like the words in this sentence.”
https://blogs.nvidia.com/blog/what-is-a-transformer-model/

Feature vector
“In pattern recognition and machine learning, a feature vector is an n-dimensional vector of numerical features that represent some object. Many algorithms in machine learning require a numerical representation of objects, since such representations facilitate processing and statistical analysis.”
https://en.wikipedia.org/wiki/Feature_(machine_learning)

Weighting
“Weights in AI refer to the numerical values that determine the strength and direction of connections between neurons in artificial neural networks. These weights are akin to synapses in biological neural networks and play a crucial role in the network’s ability to learn and make predictions.”
https://tedai-sanfrancisco.ted.com/glossary/weights/

Self-attention
“Self-attention is a mechanism used in machine learning, particularly in natural language processing (NLP) and computer vision tasks, to capture dependencies and relationships within input sequences. It allows the model to identify and weigh the importance of different parts of the input sequence by attending to itself.”
https://h2o.ai/wiki/self-attention/

Heads
“Multi-head attention is an extension of the self-attention mechanism. It enhances the model’s ability to capture diverse contextual information by simultaneously attending to different parts of the input sequence. It achieves this by performing multiple parallel self-attention operations, each with its own set of learned query, key, and value transformations.”
https://www.datacamp.com/blog/attention-mechanism-in-llms-intuition

Layers
“AI layers, also known as neural network layers, are the fundamental building blocks of artificial neural networks. They consist of three main types: input layers that receive raw data, hidden layers that process information through interconnected nodes (neurons), and output layers that produce final results.”
https://ayarlabs.com/glossary/ai-layers/