mlsanityData to 7 Oct 2026

Function-space Parameterization of Neural Networks for Sequential Learning

arXiv 2403.10929 ↗First posted 16 Mar 2024stat.ML cs.LGcited by 7

Not on the map. Why?

Abstract (arXiv)

Sequential learning paradigms pose challenges for gradient-based deep learning due to difficulties incorporating new data and retaining prior knowledge. While Gaussian processes elegantly tackle these problems, they struggle with scalability and handling rich inputs, such as images. To address these issues, we introduce a technique that converts neural networks from weight space to function space, through a dual parameterization. Our parameterization offers: (i) a way to scale function-space methods to large data sets via sparsification, (ii) retention of prior knowledge when access to past data is limited, and (iii) a mechanism to incorporate new data without retraining. Our experiments demonstrate that we can retain knowledge in continual learning and incorporate new data efficiently. We further show its strengths in uncertainty quantification and guiding exploration in model-based RL. Further information and code is available on the project website.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 46 of its 72 references that are in the map. First posted on arXiv 16 Mar 2024.

Builds on

46 of 72 references.

References · 46

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  2. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  3. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,416
  4. 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,283
  5. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,888
  6. 2015Continuous control with deep reinforcement learningcited by 3,279
  7. 2018Neural Tangent Kernel: Convergence and Generalization in Neural Networkscited by 2,597
  8. 2016Learning without Forgettingcited by 2,465
  9. 2016Variational Inference: A Review for Statisticianscited by 1,466
  10. 2018Continual Lifelong Learning with Neural Networks: A Reviewcited by 1,313
  11. 2020Dark Experience for General Continual Learning: a Strong, Simple Baselinecited by 895
  12. 2015Optimizing Neural Networks with Kronecker-factored Approximate Curvaturecited by 776
34 more
  1. 2018Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Modelscited by 748
  2. 2019Three scenarios for continual learningcited by 610
  3. 2018Progress & Compress: A scalable framework for continual learningcited by 516
  4. 2013Gaussian Processes for Big Datacited by 377
  5. 2021Laplace Redux -- Effortless Bayesian Deep Learningcited by 359
  6. 2017Variational Continual Learningcited by 335
  7. 2013(More) Efficient Reinforcement Learning via Posterior Samplingcited by 233
  8. 2019Practical Deep Learning with Bayesian Principlescited by 158
  9. 2017Practical Gauss-Newton Optimisation for Deep Learningcited by 139
  10. 2018Fast and Scalable Bayesian Deep Learning by Weight-Perturbation in Adamcited by 131
  11. 2019Functional Regularisation for Continual Learning with Gaussian Processescited by 121
  12. 2019Exact Gaussian Processes on a Million Data Pointscited by 118
  13. 2016Why is Posterior Sampling Better than Optimism for Reinforcement Learning?cited by 116
  14. 2021Scalable Marginal Likelihood Estimation for Model Selection in Deep Learningcited by 105
  15. 2020Continual Deep Learning by Functional Regularisation of Memorable Pastcited by 104
  16. 2018Measuring and regularizing networks in function spacecited by 97
  17. 2019Approximate Inference Turns Deep Networks into Gaussian Processescited by 89
  18. 2020Efficient Model-Based Reinforcement Learning through Optimistic Policy Search and Planningcited by 83
  19. 2016The Parallel Knowledge Gradient Method for Batch Bayesian Optimizationcited by 73
  20. 2020The Case for Bayesian Deep Learningcited by 68
  21. 2017Conjugate-Computation Variational Inference : Converting Variational Inference in Non-Conjugate Models to Inferences in Conjugate Modelscited by 65
  22. 2017Data-Efficient Reinforcement Learning with Probabilistic Model Predictive Controlcited by 50
  23. 2019DeepOBS: A Deep Learning Optimizer Benchmark Suitecited by 37
  24. 2023Continual Learning via Sequential Function-Space Variational Inferencecited by 28
  25. 2021GP-Tree: A Gaussian Process Classifier for Few-Shot Incremental Learningcited by 26
  26. 2020Variational Auto-Regressive Gaussian Processes for Continual Learningcited by 23
  27. 2013Model-Based Bayesian Explorationcited by 21
  28. 2021Dual Parameterization of Sparse Variational Gaussian Processescited by 19
  29. 2023Variational Linearized Laplace Approximation for Bayesian Deep Learningcited by 17
  30. 2023Posterior Sampling for Deep Reinforcement Learningcited by 16
  31. 2023Memory-Based Dual Gaussian Processes for Sequential Learningcited by 11
  32. 2022Marginal and Joint Cross-Entropies & Predictives for Online Bayesian Inference, Active Learning, and Active Samplingcited by 3
  33. 2015Sample Efficient Path Integral Control under Uncertaintycited by 2
  34. 2023Sparse Function-space Representation of Neural Networkscited by 1

Built on by

Cited by year

Citing papers per year: 2024: 4; 2025: 3.

7 papers.

  1. 2024Post-hoc Probabilistic Vision-Language Modelscited by 13
  2. 2024FSP-Laplace: Function-Space Priors for the Laplace Approximation in Bayesian Deep Learningcited by 11
  3. 2025Discrete Codebook World Models for Continuous Controlcited by 9
  4. 2024Streamlining Prediction in Bayesian Deep Learningcited by 5
  5. 2024Fixed-Mean Gaussian Processes for Post-hoc Bayesian Deep Learningcited by 2
  6. 2025Compact Memory for Continual Logistic Regressioncited by 1
  7. 2025Monte Carlo Functional Regularisation for Continual Learningcited by 0