← Search

Eliot Wong-Toi

2 accepted papers

2025

Generative Uncertainty in Diffusion Models

UAI 2025

Diffusion models have recently driven significant breakthroughs in generative modeling. While state-of-the-art models produce high-quality samples on average, individual samples can still be low quality. Detecting such samples without human inspection remains a challenging task. To address this, we

2024

Understanding Pathologies of Deep Heteroskedastic Regression

UAI 2024poster

Deep, overparameterized regression models are notorious for their tendency to overfit. This problem is exacerbated in heteroskedastic models, which predict both mean and residual noise for each data point. At one extreme, these models fit all training data perfectly, eliminating residual noise entir…

Cited by 5SourcePDFScholar