<p>Google Research just proved you can boost llm accuracy by up to 76 percentage points with zero extra output tokens, zero latency increase, and zero fine-tuning 🤯

the technique: paste your prompt twice.

that's it. that's the paper.

but WHY it works reveals something important https://t.co/Z27WXv6CGg</p>

Robert Youssef

Google Research published findings showing that simply duplicating a prompt (pasting it twice) can improve LLM accuracy by up to 76 percentage points, with no additional output tokens, no latency overhead, and no fine-tuning required. The technique works due to how LLMs process and attend to input tokens, revealing fundamental insights about how these models read and weight their context.

Google Research just proved you can boost llm accuracy by up to 76 percentage points with zero extra output tokens, zero latency increase, and zero fine-tuning 🤯

the technique: paste your prompt twice.

that's it. that's the paper.

but WHY it works reveals something important about how every llm you use actually reads your input:

<p>Google Research just proved you can boost llm accuracy by up to 76 percentage points with zero extra output tokens, zero latency increase, and zero fine-tuning 🤯

the technique: paste your prompt twice.

that's it. that's the paper.

but WHY it works reveals something important about how every llm you use actually reads your input:</p>