Which property describes LLMs that do not retain information from one prompt to the next?

Get ready for the GARP Risk and AI Exam with flashcards and multiple choice questions. Each question comes with hints and explanations. Prepare for success!

Multiple Choice

Which property describes LLMs that do not retain information from one prompt to the next?

Statelessness describes models that don’t carry memory between prompts. Each interaction is treated independently, so the model’s outputs depend only on the current input and any system prompts, not on what you sent previously. The model’s parameters remain fixed and there’s no internal memory of past exchanges unless you explicitly include that history in the current prompt or use a separate memory mechanism. This is why it best fits a behavior where information isn’t retained across prompts. Temperature affects how random or deterministic the output is, not memory. Context length (or the window) determines how much text the model can see at once in a single prompt, but it doesn’t define whether information persists across separate prompts. Transformers is the underlying architecture, not a description of memory behavior.

Statelessness describes models that don’t carry memory between prompts. Each interaction is treated independently, so the model’s outputs depend only on the current input and any system prompts, not on what you sent previously. The model’s parameters remain fixed and there’s no internal memory of past exchanges unless you explicitly include that history in the current prompt or use a separate memory mechanism. This is why it best fits a behavior where information isn’t retained across prompts.

Temperature affects how random or deterministic the output is, not memory. Context length (or the window) determines how much text the model can see at once in a single prompt, but it doesn’t define whether information persists across separate prompts. Transformers is the underlying architecture, not a description of memory behavior.

More practice questions