Preprint · 2020 · Open access

Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness

Jeremiah Zhe Liu, Zi Jun Lin, Shreyas Padhy, Dustin Tran, Tania Bedrax-Weiss, Balaji Lakshminarayanan

arXiv (Cornell University), 33, 7498–7512 · Published 17 June 2020 · DOI 10.48550/arxiv.2006.10108

Open PDFView at the publisherCite this paper

In brief

“Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness” is a 2020 preprint by Jeremiah Zhe Liu and 5 co-authors, published in arXiv (Cornell University). OpenAlex counts 686 citations as of 22 November 2026. It is open access under a CC BY licence: the full text is free to read at arXiv (Cornell University).

Citations
686
Most cited in
2025 (142)
References
85
Authors
6

Abstract

Bayesian neural networks (BNN) and deep ensembles are principled approaches to estimate the predictive uncertainty of a deep learning model. However their practicality in real-time, industrial-scale applications are limited due to their heavy memory and inference cost. This motivates us to study principled approaches to high-quality uncertainty estimation that require only a single deep neural network (DNN). By formalizing the uncertainty quantification as a minimax learning problem, we first identify input distance awareness, i.e., the model's ability to quantify the distance of a testing example from the training data in the input space, as a necessary condition for a DNN to achieve high-quality (i.e., minimax optimal) uncertainty estimation. We then propose Spectral-normalized Neural Gaussian Process (SNGP), a simple method that improves the distance-awareness ability of modern DNNs, by adding a weight normalization step during training and replacing the output layer with a Gaussian process. On a suite of vision and language understanding tasks and on modern architectures (Wide-ResNet and BERT), SNGP is competitive with deep ensembles in prediction, calibration and out-of-domain detection, and outperforms the other single-model approaches.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Liu, J. Z., Lin, Z. J., Padhy, S., Tran, D., Bedrax-Weiss, T., & Lakshminarayanan, B. (2020). Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness. arXiv (Cornell University), 33, 7498–7512. https://doi.org/10.48550/arxiv.2006.10108

MLA 9

Liu, Jeremiah Zhe, et al. “Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness.” arXiv (Cornell University), vol. 33, 2020, pp. 7498–7512. https://doi.org/10.48550/arxiv.2006.10108.

Chicago

Liu, Jeremiah Zhe, Zi Jun Lin, Shreyas Padhy, Dustin Tran, Tania Bedrax-Weiss, and Balaji Lakshminarayanan. 2020. “Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness.” arXiv (Cornell University) 33: 7498–7512. https://doi.org/10.48550/arxiv.2006.10108.

Harvard

Liu, J. Z. et al. (2020) ‘Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness’, arXiv (Cornell University), 33, pp. 7498–7512. doi:10.48550/arxiv.2006.10108.

IEEE

J. Z. Liu, Z. J. Lin, S. Padhy, D. Tran, T. Bedrax-Weiss, and B. Lakshminarayanan, “Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness,” arXiv (Cornell University), vol. 33, pp. 7498–7512, 2020, doi: 10.48550/arxiv.2006.10108.

BibTeX

@misc{liu2020simple,
  title     = {{Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness}},
  author    = {Liu, Jeremiah Zhe and Lin, Zi Jun and Padhy, Shreyas and Tran, Dustin and Bedrax-Weiss, Tania and Lakshminarayanan, Balaji},
  howpublished = {arXiv (Cornell University)},
  year      = {2020},
  volume    = {33},
  pages     = {7498--7512},
  doi       = {10.48550/arxiv.2006.10108},
  url       = {https://doi.org/10.48550/arxiv.2006.10108}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Citations by year

YearCitationsShare of the best year
201911
202077
20215757
20228888
2023134134
2024125125
2025142142
2026132132

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning

  • uncertainty estimation
  • spectral normalization
  • weight normalization
  • Bayesian neural networks
  • ensemble deep learning
  • out-of-distribution detection
  • calibration
  • visual task

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Liu, J. Z., Lin, Z. J., Padhy, S., Tran, D., Bedrax-Weiss, T., & Lakshminarayanan, B. (2020). Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness. arXiv (Cornell University), 33, 7498–7512. https://doi.org/10.48550/arxiv.2006.10108

How many times has this paper been cited?

686 times, according to OpenAlex on 22 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at arXiv (Cornell University): sign in and open the PDF from this page.

Who wrote this paper?

Jeremiah Zhe Liu, Zi Jun Lin, Shreyas Padhy, Dustin Tran, Tania Bedrax-Weiss, Balaji Lakshminarayanan.

What is the DOI of this paper?

10.48550/arxiv.2006.10108, permanently at https://doi.org/10.48550/arxiv.2006.10108. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 22 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.