What is the core concept of "Context Degradation" when using AI?	More context does not automatically mean better results; the model's attention is finite.
What is the name of the memory pattern where recall is highest for items at the beginning and end of a list?	The serial position effect (Primacy and Recency).
What phenomenon describes the drop in accuracy when key facts are placed in the center of a long context window?	"Lost in the Middle."
Why do LLMs struggle with information placed in the middle of a long context?	Transformer attention patterns naturally weight the edges (beginning and end) of the context window more heavily.
What is the primary practical advice for structuring a prompt to ensure key instructions are followed?	Place the most important instructions at the beginning and/or the end of the context.
In context engineering, what is the key difference between "more context" and "smarter context"?	"Smarter context" involves curating ruthlessly and placing information strategically, rather than just adding volume.
If a critical constraint must be followed, what is the safest prompting technique?	State the constraint early in the system prompt and restate it near the end of the context.