Long Context

You'll learn why models often ignore instructions buried in the middle of a long paste, how recall differs from use, and how to lay out prompts as Ask → Evidence → Ask again so the important rule actually counts.