Sakana AI· Sakana modelsResearch paper

String Seed of Thought: Prompting LLMs for Distribution-Faithful and Diverse Generation

Sakana AI announces a prompting technique called String Seed of Thought (SSoT) that reduces output bias and increases diversity in LLM generations, accepted at ICLR2026.

From the source

We discovered a prompting technique named String Seed of Thought (SSoT). The method is simple: instruct the LLM to generate a random string in its own output, then manipulate that string to derive its answer. It requires only a small addition to the prompt and no external random number generator.

sakana.ai

Announced

20 Apr 2026 15:00 UTC

Detected

28 Aug 2026 21:48 UTC (+130d after)