Question

A product manager at a media company is comparing two foundation models offered in Amazon Bedrock. One model has significantly more parameters than the other. The manager asks a colleague what having more parameters generally means for the model. Which statement best describes the role of parameters in a foundation model?

Accepted Answer

Parameters are the learned internal weights adjusted during training that determine how the model maps inputs to outputs; more parameters generally increase capacity but also cost and latency.

Answer

Parameters are the individual documents in the training dataset, so more parameters means the model was trained on more source data.

Answer

Parameters are the runtime settings such as temperature and top-p that a user configures to control each request.

Answer

Parameters are the maximum number of tokens the model can accept in a single prompt, so more parameters means a larger context window.