Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.09170 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908838055116800 |
|---|---|
| author | Gupta, Aditi Meyer, Raphael A. Yaniv, Yotam Chen, Elynn Erichson, N. Benjamin |
| author_facet | Gupta, Aditi Meyer, Raphael A. Yaniv, Yotam Chen, Elynn Erichson, N. Benjamin |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_09170 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Quantifying Epistemic Uncertainty in Diffusion Models Gupta, Aditi Meyer, Raphael A. Yaniv, Yotam Chen, Elynn Erichson, N. Benjamin Machine Learning Artificial Intelligence 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. |
| title | Quantifying Epistemic Uncertainty in Diffusion Models |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2602.09170 |