Which statement best describes the effect of context window size on selecting past items for inclusion?

• A. It determines how many past items are displayed to the user.
• B. It has no impact on memory usage.
• C. It requires selecting the most relevant past items for inclusion.
• D. It automatically stores all past items without cost.

Answer: (C)

A fixed context window limits how many past items the model can consider at once, so you must choose which past items to include based on their relevance to the current task. Because the window can only hold a limited amount of information, it’s not about displaying everything or avoiding cost; it’s about prioritizing the most useful past items to inform the present decision. This is why the best statement is that you must select the most relevant past items for inclusion—the window size forces a relevance-based curation of history.

Other ideas don’t fit as well because a larger window increases memory and compute, not “no impact” on memory usage; and it isn’t about automatically storing all past items without cost, since storing everything would be impractical. Also, while the window influences what can be shown, the core idea here is the necessity to pick the most pertinent past items to fit within the limited context.