← Volver

Vista previa

Editar
AWS AI Practitioner

A developer notices that an LLM produces different outputs each time it processes the same input prompt. Which characteristic of LLMs does this behavior describe?

A Nondeterministic output ✓ Correcta
B Model overfitting
C Prompt sensitivity
D Context window limitation

Explicación

LLMs are nondeterministic by nature due to the sampling process used during text generation (controlled by temperature and other parameters). The same input can produce different outputs across multiple runs because the model samples from a probability distribution of possible next tokens rather than always choosing the single most likely token.