Page 51 - CyberSecurityEssentialsEnglish
P. 51
Term Common abbreviation Plain-language explanation
A computational or logical
Model - representation that converts inputs
into an output.
Approaches based on clearly stated
Symbolic AI -
rules and knowledge.
Methods that train a model on data so
Machine learning ML
that it improves at a specific task.
A type of machine learning behind
Deep learning DL many modern image, audio, and
language applications.
Systems that create text, images,
Generative AI GenAI
audio, or other forms of content.
The stage in which a model learns from
Training -
data and examples.
Using a trained model to produce an
Inference -
output for a new case.
The information available to a model
Context -
while it constructs an answer.
An incorrect claim, number, or
Hallucination -
reference that appears convincing.
The latest period covered by the
Knowledge cutoff - model’s learned knowledge before it is
given current search or sources.
Jagged technological Uneven capability across tasks that
-
frontier may appear similar.
Realistic examples used to test a tool
Local test set -
on a specific task.
Reducing the time required for a
Acceleration -
known step.
Supporting a person’s ability to analyze
Augmentation -
and compare alternatives.
50
