TWIL: Been reading about LLM context windows & came across this article: “We find that language models often struggle to use information in the middle of long input contexts, and that performance decreases as the input context grows longer.” Research couldn’t explain why though.