A Theoretical Analysis of Memory and Overfitting Phenomena in Stochastic Interpolation Models
Two related papers explore the theoretical underpinnings of generative models, particularly focusing on stochastic interpolation. The research analyzes how these models behave with finite training data, deriving expressions for optimal fields and score functions. The findings suggest that generated samples are essentially training samples with added noise, with deviations influenced by discretization and estimation errors, leading to new definitions for overfitting and underfitting in generative contexts. AI
IMPACT Provides theoretical definitions for overfitting and underfitting in generative models, potentially guiding future research and development.