Synthetic Eggs in Many Baskets: The Impact of Synthetic Data Diversity on LLM Fine-Tuning

ArXi:2511.01490v3 Announce Type: replace As synthetic data becomes widely used in language model development, understanding its impact on model behavior is crucial. This paper investigates the impact of the diversity of sources of synthetic data on fine-tuned large language models. We focus on three key dimensions: distribution collapse, adversarial robustness, and self-preference bias. Our findings reveal that fine-tuning models on synthetic data from diverse sources can mitigate distribution collapse, preserving the breadth of the output distribution and the diversity of the output text.