Context windows and memory

LLMs have a context window limiting how much text they process at once. Techniques like sliding windows and summarization extend effective memory.

• Context windows limit how much text models see.
• Techniques exist to extend effective memory.