Prompting & context · Glossary term
What is Lost in the Middle?
A long-context failure pattern in which model performance changes with evidence position and can degrade when relevant information sits between the beginning and end.
Why does Lost in the Middle matter?
Fitting evidence inside the context window does not guarantee that the model will use every position with equal reliability.
Lost in the Middle in practice
Test several evidence positions, reduce distractors, place decision-critical constraints where they remain salient, and verify answers against the source.
What is the common confusion about Lost in the Middle?
It is an observed behavior pattern, not a fixed law that affects every model, task, or position identically.
Learn Lost in the Middle in the course
No lesson links to this term yet. Search the course catalog for it.
Related terms
- Context WindowThe maximum token capacity available to one model inference under a specific model and API contract.
- Context EngineeringDesigning the full information environment supplied to a model at each step, including instructions, selected files, retrieved evidence,…
- Eval SetA versioned collection of inputs, expected properties, scoring rules, and metadata used to measure an AI system against a defined…
- GroundingConnecting a generated answer or action to evidence, state, or observations that the system can identify and check.
Sources
More terms in Prompting & context
This entry comes from glossary/terms.md on GitHub. Browse all 250 glossary terms.