Research
String Seed of Thought: Prompting LLMs for Distribution-Faithful and Diverse Generation https://arxiv.org/abs/2510.21150 https://pub.sakana.…
This paper presents a prompting technique called 'String Seed of Thought' that enables large language models to generate outputs that are both faithful to underlying data distributions and maintain di
This paper presents a prompting technique called "String Seed of Thought" that enables large language models to generate outputs that are both faithful to underlying data distributions and maintain diversity in their responses. The method addresses the challenge of balancing distribution accuracy with generation variety when using LLMs. The work is associated with Sakana AI and appears to focus on improving how LLMs sample from learned distributions during generation.
Related
- Getting LLMs to simulate “true” randomness or generate diverse outputs is surprisingly difficult. We found a simple prompting trick that sol…
- Private Seeds, Public LLMs: Realistic and Privacy-Preserving Synthetic Data Generation
- CircuitSynth: Reliable Synthetic Data Generation
- Adaptive Conformal Prediction for Improving Factuality of Generations by Large Language Models
Source: David Ha (X) | 2026-04-21