Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash 89 tok/s
Gemini 2.5 Pro 54 tok/s Pro
GPT-5 Medium 27 tok/s
GPT-5 High 22 tok/s Pro
GPT-4o 89 tok/s
GPT OSS 120B 457 tok/s Pro
Kimi K2 169 tok/s Pro
2000 character limit reached

Layer Ensembles (2210.04882v3)

Published 10 Oct 2022 in cs.LG

Abstract: Deep Ensembles, as a type of Bayesian Neural Networks, can be used to estimate uncertainty on the prediction of multiple neural networks by collecting votes from each network and computing the difference in those predictions. In this paper, we introduce a method for uncertainty estimation that considers a set of independent categorical distributions for each layer of the network, giving many more possible samples with overlapped layers than in the regular Deep Ensembles. We further introduce an optimized inference procedure that reuses common layer outputs, achieving up to 19x speed up and reducing memory usage quadratically. We also show that the method can be further improved by ranking samples, resulting in models that require less memory and time to run while achieving higher uncertainty quality than Deep Ensembles.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (19)
  1. David J C Mackay, “Probable networks and plausible predictions — a review of practical bayesian methods for supervised neural networks,” Network, vol. 6, no. 3, pp. 469–505, 1995.
  2. “Bayesian Deep Learning and a Probabilistic Perspective of Generalization,” in NeurIPS, 2020, vol. 33.
  3. “Bayesian neural networks,” arXiv:2006.01490, 2020.
  4. “Bayesian learning for neural networks: an algorithmic survey,” arxiv:2211.11865, 2022.
  5. “Weight Uncertainty in Neural Networks,” in ICML, 2015, vol. 37, pp. 1613–1622.
  6. “Hypermodels for exploration,” in ICLR, 2020, vol. 8.
  7. “Dropout as a bayesian approximation: Representing model uncertainty in deep learning,” in ICML, 2016, vol. 48, pp. 1050–1059.
  8. “Randomized prior functions for deep reinforcement learning,” in NeurIPS, 2018, vol. 31, pp. 8626–8638.
  9. “Epistemic Neural Networks,” arXiv:2107.08924, 2021.
  10. “Deep neural networks as gaussian processes,” in ICLR, 2018.
  11. S. Kullback and R. A. Leibler, “On Information and Sufficiency,” Ann. Math. Stat., vol. 22, no. 1, pp. 79 – 86, 1951.
  12. “Dropout: A simple way to prevent neural networks from overfitting,” JMLR, vol. 15, no. 56, pp. 1929–1958, 2014.
  13. “Auto-encoding variational bayes,” in ICLR, 2014, vol. 2.
  14. “Variational neural networks,” arxiv:2207.01524, 2022.
  15. Matias Valdenegro-Toro, “Deep sub-ensembles for fast uncertainty estimation in image classification,” arxiv:1910.08168, 2019.
  16. “Batchensemble: An alternative approach to efficient ensemble and lifelong learning,” in ICLR, 2020, vol. 8.
  17. “Layer ensembles: A single-pass uncertainty estimation in deep learning for segmentation,” in MICCAI, 2022, vol. 13438, pp. 514–524.
  18. “Conditional deep learning for energy-efficient and enhanced pattern recognition,” in DATE, 2016, pp. 475–480.
  19. “Github code for epistemic neural networks,” https://github.com/deepmind/enn, 2021.
Citations (1)
List To Do Tasks Checklist Streamline Icon: https://streamlinehq.com

Collections

Sign up for free to add this paper to one or more collections.

Summary

We haven't generated a summary for this paper yet.

Ai Generate Text Spark Streamline Icon: https://streamlinehq.com

Paper Prompts

Sign up for free to create and run prompts on this paper using GPT-5.

Dice Question Streamline Icon: https://streamlinehq.com

Follow-up Questions

We haven't generated follow-up questions for this paper yet.