# The Hidden Uncertainty in a Neural Networks Activations

arXiv: https://arxiv.org/abs/2012.03082 · HTML: https://mlsanity.com/paper/2012.03082/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Dec 2020.

## Built on by (most cited first, 14 of 14)

- [Deep Deterministic Uncertainty: A Simple Baseline](https://mlsanity.com/paper/2102.11582/) (2021)
- [Rethinking Aleatoric and Epistemic Uncertainty](https://mlsanity.com/paper/2412.20892/) (2024)
- [Beyond Confidence: Reliable Models Should Also Consider Atypicality](https://mlsanity.com/paper/2305.18262/) (2023)
- [Deep Bayesian Active Learning for Preference Modeling in Large Language Models](https://mlsanity.com/paper/2406.10023/) (2024)
- [Hyperdimensional Feature Fusion for Out-Of-Distribution Detection](https://mlsanity.com/paper/2112.05341/) (2021)
- [Gaussian Latent Representations for Uncertainty Estimation using Mahalanobis Distance in Deep Classifiers](https://mlsanity.com/paper/2305.13849/) (2023)
- [Generating and Detecting True Ambiguity: A Forgotten Danger in DNN Supervision Testing](https://mlsanity.com/paper/2207.10495/) (2022)
- [Function Space Diversity for Uncertainty Prediction via Repulsive Last-Layer Ensembles](https://mlsanity.com/paper/2412.15758/) (2024)
- [Training, Architecture, and Prior for Deterministic Uncertainty Methods](https://mlsanity.com/paper/2303.05796/) (2023)
- [Discriminant Distance-Aware Representation on Deterministic Uncertainty Quantification Methods](https://mlsanity.com/paper/2402.12664/) (2024)
- [Generative Adversarial Models for Extreme Geospatial Downscaling](https://mlsanity.com/paper/2402.14049/) (2024)
- [When to ASK: Uncertainty-Gated Language Assistance for Reinforcement Learning](https://mlsanity.com/paper/2604.02226/) (2026)
- [Can a single neuron learn predictive uncertainty?](https://mlsanity.com/paper/2106.03702/) (2021)
- [Uncertainty Estimation in Instance Segmentation of Affordances via Bayesian Visual Transformers](https://mlsanity.com/paper/2605.03614/) (2026)

## Abstract (arXiv)

The distribution of a neural network's latent representations has been successfully used to detect out-of-distribution (OOD) data. This work investigates whether this distribution moreover correlates with a model's epistemic uncertainty, thus indicates its ability to generalise to novel inputs. We first empirically verify that epistemic uncertainty can be identified with the surprise, thus the negative log-likelihood, of observing a particular latent representation. Moreover, we demonstrate that the output-conditional distribution of hidden representations also allows quantifying aleatoric uncertainty via the entropy of the predictive distribution. We analyse epistemic and aleatoric uncertainty inferred from the representations of different layers and conclude that deeper layers lead to uncertainty with similar behaviour as established - but computationally more expensive - methods (e.g. deep ensembles). While our approach does not require modifying the training process, we follow prior work and experiment with an additional regularising loss that increases the information in the latent representations. We find that this leads to improved OOD detection of epistemic uncertainty at the cost of ambiguous calibration close to the data distribution. We verify our findings on both classification and regression models.