5 papers · 1 filter
Recursive Learning Without Collapse: A Weighting-Based Stabilization Framework
Hengzhi He, Shirong Xu, Guang Cheng
Recent studies identified an intriguing phenomenon in recursive generative model training known as model collapse, where models trained on data generated by previous models exhibit…
A Probabilistic Perspective on Model Collapse
Shirong Xu, Hengzhi He, Guang Cheng
In recent years, model collapse has become a critical issue in language model training, making it essential to understand the underlying mechanisms driving this phenomenon. In this…
Utility Theory of Synthetic Data Generation
Shirong Xu, Will Wei Sun, Guang Cheng
Synthetic data algorithms are widely employed in industries to generate artificial data for downstream learning tasks. While existing research primarily focuses on empirically eval…
Rate-Optimal Rank Aggregation with Private Pairwise Rankings
Shirong Xu, Will Wei Sun, Guang Cheng
In various real-world scenarios, such as recommender systems and political surveys, pairwise rankings are commonly collected and utilized for rank aggregation to derive an overall…
Discriminative Estimation of Total Variation Distance: A Fidelity Auditor for Generative Data
Lan Tao, Shirong Xu, Chi-Hua Wang +2
With the proliferation of generative AI and the increasing volume of generative data (also called as synthetic data), assessing the fidelity of generative data has become a critica…