Quantifying Epistemic Uncertainty in Diffusion Models
This work addresses uncertainty estimation for diffusion model users, but it is incremental as it builds on existing methods with a scalable approximation.
The paper tackled the problem of unreliable epistemic uncertainty quantification in diffusion models by introducing a method based on Fisher information that isolates epistemic variance, resulting in more accurate and reliable filtering in synthetic time-series generation tasks.
To ensure high quality outputs, it is important to quantify the epistemic uncertainty of diffusion models.Existing methods are often unreliable because they mix epistemic and aleatoric uncertainty. We introduce a method based on Fisher information that explicitly isolates epistemic variance, producing more reliable plausibility scores for generated data. To make this approach scalable, we propose FLARE (Fisher-Laplace Randomized Estimator), which approximates the Fisher information using a uniformly random subset of model parameters. Empirically, FLARE improves uncertainty estimation in synthetic time-series generation tasks, achieving more accurate and reliable filtering than other methods. Theoretically, we bound the convergence rate of our randomized approximation and provide analytic and empirical evidence that last-layer Laplace approximations are insufficient for this task.