Glossary

A practical field guide to AI, ML, data science, mathematics, and CS terms.

A crawlable glossary built for readers who want intuition first, technical precision second, and enough context to connect one concept to the next.

Machine Learning Foundations Deep Learning Transformers and LLMs AI Systems Data Science and Statistics Mathematics for ML Computer Science

Transformers and LLMs

Attention

Attention is a neural network mechanism that lets a model decide which parts of an input are most relevant when producing a representation or output.

Read definition Transformers and LLMs

Transformer

A transformer is a neural network architecture built around attention mechanisms and feed-forward layers.

Read definition AI Systems

Retrieval-Augmented Generation

Retrieval-augmented generation combines search or retrieval with a generative model so answers can use external context.

Read definition Deep Learning

Embedding

An embedding is a learned vector representation of an object such as a word, document, image, user, or product.

Read definition Machine Learning Foundations

Gradient Descent

Gradient descent is an optimization method that updates model parameters in the direction that most reduces the loss locally.

Read definition AI Systems

Mixture of Experts

Mixture of Experts is a model architecture that routes inputs to a subset of specialized expert networks instead of activating every parameter.

Read definition

Index

Browse the glossary by technical area.

Each term page includes a plain-English definition, intuition, a technical definition, example, related terms, and sources.

Deep Learning

Neural network building blocks, representations, activations, and regularization ideas.

10 terms

Activation FunctionAlso: nonlinearity Batch NormalizationAlso: batch norm ConvolutionAlso: conv Deep LearningAlso: DL EmbeddingAlso: vector embedding Latent SpaceAlso: representation space Neural NetworkAlso: artificial neural network ReLUAlso: Rectified Linear Unit SigmoidAlso: logistic function Softmax

Transformers and LLMs

Attention, tokens, context, multimodal architecture, and inference controls for modern language models.

17 terms

AttentionAlso: attention mechanism Context WindowAlso: context length Cross-Attention Hybrid Attention KV CacheAlso: key-value cache Multimodal ModelAlso: multimodal AI Positional EncodingAlso: position embedding Reasoning EffortAlso: effort setting, thinking budget Relative AttentionAlso: relative position attention RoPEAlso: Rotary Position Embedding, rotary embedding Self-Attention Short ConvolutionAlso: SConv, short 1D convolution Sliding-Window AttentionAlso: local attention TemperatureAlso: sampling temperature Thin Modality TowersAlso: light modality encoders, thin modality adapters Token Transformer

AI Systems

Production patterns for retrieval, agents, tool use, evaluation, guardrails, and serving.

16 terms

Active Parameters AgentAlso: AI agent Fine-TuningAlso: finetuning GuardrailAlso: AI guardrail, safety control HallucinationAlso: model hallucination InferenceAlso: model serving Instruction Tuning LatencyAlso: response time Mixture of ExpertsAlso: MoE Model EvaluationAlso: AI evaluation, evals Prompt EngineeringAlso: prompt design Retrieval-Augmented GenerationAlso: RAG Semantic Search Sparse RoutingAlso: expert routing Tool UseAlso: function calling Vector DatabaseAlso: vector DB

Data Science and Statistics

Metrics, uncertainty, causal thinking, and statistical concepts used to reason from data.

13 terms

Accuracy AUC ROCAlso: ROC AUC, area under the ROC curve CalibrationAlso: probability calibration CausationAlso: causal effect Confidence IntervalAlso: CI Confusion Matrix Correlation Distribution ShiftAlso: data drift F1 ScoreAlso: F-score, F measure P-ValueAlso: p value PrecisionAlso: positive predictive value RecallAlso: sensitivity, true positive rate Statistical Significance

Mathematics for ML

Linear algebra, probability, optimization, and geometry concepts behind model behavior.

13 terms

Computer Science

Algorithms, systems, concurrency, distributed systems, and reliability terms that shape production software.

16 terms

Algorithm Byzantine Fault ToleranceAlso: BFT Cache Cache Coherence CAP TheoremAlso: Brewer theorem Concurrency Consensus Deadlock Distributed System Hash TableAlso: hash map IdempotencyAlso: idempotent operation IndexAlso: data index MutexAlso: mutual exclusion lock Race ConditionAlso: data race Space ComplexityAlso: memory complexity Time ComplexityAlso: runtime complexity

Start here

Want these definitions connected to deeper AI explainers?

The glossary is now the internal reference layer. Next, we can link new and existing articles into these definitions so readers and chatbot engines can follow the concept graph.

Read the Journal