Home Contents Glossary Test Yourself About

83 terms

Glossary

Plain-English definitions of AI engineering terms: what people say, what it actually means, and why it's called that.

A

Activation Function
Adam (Optimizer)
AdamW
Agent
Alignment
Attention
Autograd
Autoregressive

B

Backpropagation
Batch Size

C

Chain of Thought (CoT)
Chunking
CNN (Convolutional Neural Network)
Context Window
Contrastive Learning
Cosine Similarity
Cross-Entropy
CUDA

D

Data Augmentation
Decoder
Diffusion Model
DPO (Direct Preference Optimization)
Dropout

E

Eigenvalue
Embedding
Encoder
Epoch

F

Feature
Few-Shot
Fine-tuning
Function Calling

G

GAN (Generative Adversarial Network)
GPT
Gradient
Gradient Descent
Guardrails

H

Hallucination
Hyperparameter

I

Inductive Bias
Inference

J

JAX

K

KV Cache

L

Latent Space
Learning Rate
LLM (Large Language Model)
LoRA (Low-Rank Adaptation)
Loss Function

M

MCP (Model Context Protocol)
Mixed Precision
MoE (Mixture of Experts)

N

NaN (Not a Number)
Normalization

O

Optimizer
Overfitting

P

Parameter
Perplexity
Precision & Recall
Prompt Engineering
Prompt Injection

Q

QLoRA
Quantization

R

RAG (Retrieval-Augmented Generation)
ReLU
RLHF (Reinforcement Learning from Human Feedback)
ROUGE

S

Self-Attention
Semantic Search
SFT (Supervised Fine-Tuning)
Softmax
Streaming
Swarm
System Prompt

T

Temperature
Tensor
Token
Transfer Learning
Transformer

U

Underfitting

V

VAE (Variational Autoencoder)
Vector Database

W

Weight
Weight Decay

Z

Zero-Shot

© 2026 Tai Bui