Agentic LoopThe cycle where an AI model takes an action, observes the result, then decides what to do next.3 Apr 2026agentsarchitecturereasoning
Attention MechanismThe core operation inside a transformer model that lets it weigh how much each word relates to every other word.3 Apr 2026architecturemodelsconcept
Context WindowThe maximum amount of text a model can hold in its working memory at once.3 Apr 2026architecturemodelsconcept
Knowledge GraphA structured representation of facts as entities and relationships.3 Apr 2026architectureragconcept
Large Language Model (LLM)A neural network trained to predict and generate text at scale.3 Apr 2026modelsconceptarchitecture
Mixture of Experts (MoE)An architecture where only a subset of the model's parameters activate for each input.3 Apr 2026architecturemodelsperformance
Multi-ModalA model that handles more than one type of input or output, like text plus images.3 Apr 2026modelsarchitectureconcept
Retrieval-Augmented Generation (RAG)A technique that grounds LLM responses in external data by retrieving relevant documents before generating a reply.3 Apr 2026architectureretrievalllmgrounding
Sparse AttentionA transformer efficiency technique where each token attends only to a relevant subset of other tokens, enabling near-linear scaling with context length.3 Apr 2026architecturetransformersefficiencylong-contextresearch
TokenisationThe process of splitting text into tokens, the chunks a model actually processes.3 Apr 2026conceptmodelsarchitecture
Tool UseThe ability of a model to call external functions or APIs to complete a task.3 Apr 2026agentstoolingarchitecture