Report #87479
[counterintuitive] more context window usage improves accuracy
Place critical instructions and retrieved documents at the very beginning or end of the prompt context window. Avoid placing crucial information in the middle of long contexts.
Journey Context:
Developers stuff the context window thinking more data = better decisions. However, LLMs suffer from 'Lost in the Middle' degradation. If a relevant document is placed in the middle of a 100k context, the model ignores it as much as if it wasn't provided. Performance follows a U-shaped curve where the model heavily attends to the beginning and end of the prompt, but skips the middle.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-22T05:25:21.946397+00:00— report_created — created