Search the practice set

126 questions, 454 terms and 400 topics in 20 areas.

18 results for “LLM Evaluation”

Terminology · 12
term

LLM Evaluation

A repeatable measurement of whether an AI system's outputs are good enough, on cases that reflect the actual task.

AI-Era Architecture
term

Prompt Registry

A versioned store of production prompts with their model bindings, parameters and evaluation results, so a prompt change is a reviewable, traceable, reversible deployment.

AI-Era Architecture
term

Prompt Versioning

Treating prompts as versioned, reviewed, tested artefacts rather than as strings edited in place.

AI-Era Architecture
term

AI Gateway

A shared proxy in front of model providers that centralises routing, keys, quotas, caching, logging and safety policy.

AI-Era Architecture
term

Architecture Trade-off Analysis Method

A structured evaluation that scores an architecture against prioritised quality-attribute scenarios and identifies the points where those attributes conflict.

Architecture Decision-Making
term

Context Window

The maximum number of tokens a model can attend to in one request, holding the system prompt, history, retrieved context, tools and the answer.

AI-Era Architecture
term

Guardrail

A deterministic check applied to a model's input or output, enforcing rules that cannot be left to the model itself.

AI-Era Architecture
term

Model Router

Directing each request to a model chosen by the task's difficulty, cost and latency budget, rather than sending everything to the largest model available.

AI-Era Architecture
term

Prompt Injection

An attack in which text from an untrusted source is interpreted by the model as instructions rather than as data.

AI-Era Architecture
term

Retrieval-Augmented Generation

Retrieving relevant documents at query time and putting them in the model's context, so answers are grounded in your data rather than in training data.

AI-Era Architecture
term

Semantic Cache

Caching model responses keyed by the meaning of the request rather than by its exact text, so near-duplicate questions are served without a model call.

AI-Era Architecture
term

Tool Calling

Giving a model a set of typed function definitions it can request to invoke, with the application executing the call and returning the result.

AI-Era Architecture