mlsanityData to 14 Oct 2026

Task Agnostic Continual Learning Using Online Variational Bayes

arXiv 1803.10123 ↗First posted 27 Mar 2018stat.ML cs.LGcited by 61

Not on the map. Why?

Abstract (arXiv)

Catastrophic forgetting is the notorious vulnerability of neural networks to the change of the data distribution while learning. This phenomenon has long been considered a major obstacle for allowing the use of learning agents in realistic continual learning settings. A large body of continual learning research assumes that task boundaries are known during training. However, research for scenarios in which task boundaries are unknown during training has been lacking. In this paper we present, for the first time, a method for preventing catastrophic forgetting (BGD) for scenarios with task boundaries that are unknown during training --- task-agnostic continual learning. Code of our algorithm is available at https://github.com/igolan/bgd.

Key facts. Cited by 61 papers in the map (data through 14 Oct 2026). First posted on arXiv 27 Mar 2018.

Builds on

No reference list yet.

Built on by

Cited by year

Citing papers per year: 2021: 14; 2022: 17; 2023: 14; 2024: 14; 2025: 2.

61 papers.

Sort:

By area:

  1. 2021Learning to Prompt for Continual Learningcited by 944
  2. 2021New Insights on Reducing Abrupt Representation Change in Online Continual Learningcited by 219
  3. 2023RanPAC: Random Projections and Pre-trained Models for Continual Learningcited by 164
  4. 2022The CLEAR Benchmark: Continual LEArning on Real-World Imagerycited by 93
  5. 2024Continual Learning and Catastrophic Forgettingcited by 77
  6. 2021Continual Learning via Local Module Compositioncited by 63
  7. 2023Computationally Budgeted Continual Learning: What Does Matter?cited by 61
  8. 2022A Closer Look at Rehearsal-Free Continual Learningcited by 54
  9. 2024Self-Expansion of Pre-trained Models with Mixture of Adapters for Continual Learningcited by 53
  10. 2021Layerwise Optimization by Gradient Decomposition for Continual Learningcited by 51
  11. 2021Posterior Meta-Replay for Continual Learningcited by 45
  12. 2023Traj-MAE: Masked Autoencoders for Trajectory Predictioncited by 42
38 more
  1. 2021Class-Incremental Learning with Generative Classifierscited by 39
  2. 2021Continual Learning for Recurrent Neural Networks: an Empirical Evaluationcited by 34
  3. 2022Improving Task-free Continual Learning by Distributionally Robust Memory Evolutioncited by 31
  4. 2022Learning Bayesian Sparse Networks with Full Experience Replay for Continual Learningcited by 29
  5. 2024Online-LoRA: Task-free Online Continual Learning via Low Rank Adaptationcited by 27
  6. 2022Reliable amortized variational inference with physics-based latent distribution correctioncited by 25
  7. 2022Challenging Common Assumptions about Catastrophic Forgettingcited by 20
  8. 2021Sequoia: A Software Framework to Unify Continual Learning Researchcited by 19
  9. 2021Encoders and Ensembles for Task-Free Continual Learningcited by 17
  10. 2021Task-agnostic Continual Learning with Hybrid Probabilistic Modelscited by 16
  11. 2023Knowledge Accumulation in Continually Learned Representations and the Issue of Feature Forgettingcited by 16
  12. 2023Prediction Error-based Classification for Class-Incremental Learningcited by 14
  13. 2021Towards Visual Semanticscited by 12
  14. 2021ACAE-REMIND for Online Continual Learning with Compressed Feature Replaycited by 12
  15. 2021Towards a theory of out-of-distribution learningcited by 11
  16. 2022Task-Agnostic Continual Reinforcement Learning: Gaining Insights and Overcoming Challengescited by 11
  17. 2022Beyond Supervised Continual Learning: a Reviewcited by 11
  18. 2024Random Representations Outperform Online Continually Learned Representationscited by 11
  19. 2022TAME: Task Agnostic Continual Learning using Multiple Expertscited by 8
  20. 2023Clustering-based Domain-Incremental Learningcited by 8
  21. 2024Continual Learning with Neuromorphic Computing: Foundations, Methods, and Emerging Applicationscited by 8
  22. 2022Self-Activating Neural Ensembles for Continual Reinforcement Learningcited by 7
  23. 2023On Sequential Bayesian Inference for Continual Learningcited by 6
  24. 2025Knowledge Adaptation as Posterior Correctioncited by 6
  25. 2022Forgetting and Imbalance in Robot Lifelong Learning with Off-policy Datacited by 5
  26. 2023Approximate Bayesian Class-Conditional Models under Continuous Representation Shiftcited by 5
  27. 2023Online Continual Learning via Logit Adjusted Softmaxcited by 5
  28. 2024Prospective Learning: Learning for a Dynamic Futurecited by 4
  29. 2022A Study of Continual Learning Methods for Q-Learningcited by 3
  30. 2022Integral Continual Learning Along the Tangent Vector Field of Taskscited by 3
  31. 2022Instance-Conditional Timescales of Decay for Non-Stationary Learningcited by 3
  32. 2023Adiabatic replay for continual learningcited by 3
  33. 2024Self-Regulated Neurogenesis for Online Data-Incremental Learningcited by 3
  34. 2023One-Shot Machine Unlearning with Mnemonic Codecited by 2
  35. 2021Continual Learning with Neuron Activation Importancecited by 1
  36. 2022DIRA: Dynamic Domain Incremental Regularised Adaptationcited by 1
  37. 2022Prototypical quadruplet for few-shot class incremental learningcited by 1
  38. 2023Algorithm Design for Online Meta-Learning with Task Boundary Detectioncited by 1

50 most cited of 61 All 61 (JSON)