Skip to content

SOP-005: Lost in the Middle

Fresh Updated January 2026
Document Control
SOP IDSOP-005
Version1.0
StatusActive
SourceChroma Research (2025)

Overview

The "Lost in the Middle" phenomenon describes how LLMs disproportionately attend to information at the beginning and end of their context window, while systematically ignoring content in the middle.

The Position Effect

Key Research Data

Position vs Performance

PositionRetrieval AccuracyRelative Performance
First 20%~100%Baseline
Middle 40-60%~70%-30%
Last 20%~80-90%-10 to -20%

Query Position Matters

Needle PositionAccuracyNotes
Unique token early100%Best case
Unique token middle20-50%Severe degradation
Unique token late60-80%Partial recovery

Critical

Buried content loses 80% citability at scale.

Why This Happens

Attention is unevenly distributed - even on simple retrieval tasks, middle positions are systematically ignored.

Benchmark vs Reality Gap

BenchmarkResultReal-World Performance
NIAH (Needle in Haystack)80-100%N/A - synthetic
NoLiMa (Semantic matching)40-70%More realistic
LongMemEval (Conversational)50-80%Most realistic

WARNING

Long-context benchmarks are misleading. Real use fails silently because models appear to work on synthetic tests.

Content Strategy Implications

Do: Front-Load Critical Information

Content Positioning Guidelines

Content TypeRecommended PositionWhy
Primary claimsFirst 20%Maximum attention
Key differentiatorsFirst 20%Citation probability
Supporting evidenceMiddle 60%Less critical
Call to actionLast 20%Recency helps
Summary/repeatLast 20%Reinforcement

Practical Implementation

For Blog Posts / Articles

[First 20%]
- State main thesis immediately
- Include key claims upfront
- Front-load unique data points

[Middle 60%]
- Supporting arguments
- Examples and evidence
- Background information

[Last 20%]
- Summarize key points
- Repeat most important claims
- Call to action

For Product Pages

[First 20%]
- Product name + primary benefit
- Key differentiator
- Most important feature

[Middle 60%]
- Feature details
- Technical specifications
- Comparisons

[Last 20%]
- Repeat primary benefit
- Social proof summary
- Clear CTA

Verification Checklist

  • [ ] Audit existing content for buried key points
  • [ ] Move critical claims to first 20% of content
  • [ ] Repeat key points in last 20%
  • [ ] Test content retrieval with position-specific queries
  • [ ] Structure new content with position awareness

A/B Testing Approach

See Also

Citations

"Models perform best when target info at start/end; middle positions ~30% worse" — Chroma Context Rot Literature

"Position accuracy: 100% when unique token early → <20% when late" — Chroma Haystack Position Experiment

Based on research from Thinking Machines Lab, Chroma Research, and ACL 2024-2025