Why Models Behave Differently: Parameters, Variability, and Control
Participants hold a task steady, vary one condition and repeat trials. They use the evidence to decide when variation helps, when it creates risk and which checks are needed.
AI fundamental
Model behaviour and variability
Tokens, tokenisation, attention, sampling, parameters, repeated runs, and observable variability.
Guiding question
Which generation condition changed, and is the observed difference stable enough to interpret?
Use observed results to support the explanation. State what the experiment cannot establish.
Related units
Where this fundamental is taught
Open the complete Friday guide for the conceptual input, demonstration, experiment, caveat, and transfer activity.