The Verbalized Sampling Prompt

A Stanford research-backed technique that boosts LLM creativity by 1.6-2x. Add ~20 words to any prompt to restore the diversity lost after RLHF alignment.

Modern LLMs are less creative than they could be. Post-training alignment (RLHF) makes them helpful and safe — but it also causes mode collapse, where the model favors a narrow set of predictable responses over diverse, creative ones.

Verbalized Sampling is a training-free prompting technique that bypasses this limitation. No fine-tuning required — just change how you ask.

The Results

1.6-2x
Creativity boost over direct prompting
25.7%
Higher human-rated diversity
66.8%
Lost creativity restored
0
Training required

The Technique

Direct Prompting
Tell me a joke.
Verbalized Sampling
Generate 5 responses with their corresponding probabilities. Tell me a joke.

That's it. By asking for a distribution instead of a single instance, you force the model to tap into the diverse knowledge it learned during pre-training — before alignment narrowed its outputs.

Why This Works

The Problem: Typicality Bias

During RLHF, human annotators rate LLM responses. But humans naturally prefer answers that are familiar, easy to read, and predictable — even when creative alternatives are equally good. This "typicality bias" gets baked into the reward model, which then aggressively sharpens the LLM's probability distribution toward safe, boring outputs.

How Verbalized Sampling Bypasses Mode Collapse
1
Direct prompt → Aligned personality activates → Most reinforced answer
2
VS prompt → Model asked for distribution → Must access full knowledge
3
Pre-training weights activated → Diverse responses surface

Key insight: The LLM still has two personalities after alignment — the original pre-trained model with rich, diverse knowledge, and the safety-focused aligned model. Verbalized Sampling acts as a "mental switch" to access the original.

Variants

VS-CoT
Verbalized Sampling + Chain-of-Thought for complex reasoning with diversity
VS-Multi
Request multiple diverse outputs in a single generation

When to Use This

  • Creative writing — stories, jokes, metaphors, analogies
  • Brainstorming — idea generation, product names, taglines
  • Problem solving — when you want multiple approaches, not just the obvious one
  • Content variety — social posts, email variants, headlines
  • Any task where "predictable" = bad

Example Applications

Creative Writing

"Generate 5 responses with their corresponding probabilities. Write a one-sentence horror story."

Brainstorming

"Generate 5 responses with their corresponding probabilities. Give me a startup idea for the education space."

Marketing Copy

"Generate 5 responses with their corresponding probabilities. Write a tagline for a productivity app."

Stanford Research

This technique comes from Stanford researchers studying mode collapse in aligned LLMs. The paper demonstrates that verbalized sampling significantly enhances diversity (1.6–2.1x) while maintaining or improving quality.

The Template

Generate 5 responses with their corresponding probabilities. [Your actual prompt here]

Just prepend those 8 words to any prompt. The model will output multiple options with probability estimates, giving you access to its full creative range instead of just the most "aligned" answer.