Valid Inference with Synthetic Data via Task Exchangeability

There is a proliferation of work arguing for the use of synthetic data in scientific research. For example, social scientists are arguing for the use of LLM-generated "silicon samples" in pilot studies; AI evaluations increasingly rely on "LLM-as-a-judge" outputs; and proteomics research is accelerated by generative models that produce synthetic protein structures. These developments raise an intriguing possibility: synthetic data may help researchers ask more questions, run more studies, and accelerate discovery. But they also raise a fundamental concern: synthetic data can be biased, noisy, and misspecified. In this work, we propose statistical principles for using synthetic data in scientific research with provable validity guarantees. The key insight is a new technical condition that w

This Coalition transmission preserves metadata, trust context, and source routing for a Research Wire item from arXiv. The complete article remains with the original publisher.

Content handling is marked as Metadata Only. Use the source article link for the full report, updates, corrections, and publisher-controlled context.