Reexamining the Aleatoric and Epistemic Uncertainty Dichotomy

Michael Kirchhof, Gjergji Kasneci, Enkelejda Kasneci

Blogpost Url: https://d2jud02ci9yv69.cloudfront.net/2025-04-28-reexamining-the-aleatoric-and-epistemic-uncertainty-dichotomy-93/blog/reexamining-the-aleatoric-and-epistemic-uncertainty-dichotomy/

Abstract:

When discussing uncertainty estimates for the safe deployment of AI agents in the real world, the field typically distinguishes between aleatoric and epistemic uncertainty. This dichotomy may seem intuitive and well-defined at first glance, but this blog post reviews examples, quantitative findings, and theoretical arguments that reveal that popular definitions of aleatoric and epistemic uncertainties directly contradict each other and are intertwined in fine nuances. We peek beyond the epistemic and aleatoric uncertainty dichotomy and reveal a spectrum of uncertainties that help solve practical tasks especially in the age of large language models.

Conflict Of Interest:

The authors of this blog post occur on only 2 of the 36 cited papers. By presenting each of the (partially conflicting) perspectives on aleatoric and epistemic uncertainty side by side without taking a position, we strive to ensure neutrality and provide the reader with an unbiased review of the field. We are happy to address any issues that should be brought up during the review to make this a neutral, trustworthy perspective on the field of uncertainty decomposition.

Submission Number: 58