Development

Repeating a prompt twice improves performance for free

development2026-01-11

Claim

Google researchers found that simply repeating a prompt twice lets models attend to their own inputs without reasoning, improving performance at zero generated-token cost.

Reporting date: 2026-01-11 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to January 11, 2026 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

Reported in

Welcome to January 11, 2026

Actor

Google

Evidence for

Referenced by

Welcome to January 11, 2026

Cite or correct this record

Citation formatCitation text

Repeating a prompt twice improves performance for free
Innermost Loop KB — secondary record
Reporting date: 2026-01-11
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-01-11-repeat-the-prompt-twice
Claim: Google researchers found that simply repeating a prompt twice lets models attend to their own inputs without reasoning, improving performance at zero generated-token cost.
Attributed sources:
- Welcome to January 11, 2026 — https://theinnermostloop.substack.com/p/welcome-to-january-11-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.