mlsanityData to 14 Oct 2026

Hebbian Learning based Orthogonal Projection for Continual Learning of Spiking Neural Networks

arXiv 2402.11984 ↗First posted 19 Feb 2024cs.NE cs.AI cs.LGcited by 8

Not on the map. Why?

Abstract (arXiv)

Neuromorphic computing with spiking neural networks is promising for energy-efficient artificial intelligence (AI) applications. However, different from humans who continually learn different tasks in a lifetime, neural network models suffer from catastrophic forgetting. How could neuronal operations solve this problem is an important question for AI and neuroscience. Many previous studies draw inspiration from observed neuroscience phenomena and propose episodic replay or synaptic metaplasticity, but they are not guaranteed to explicitly preserve knowledge for neuron populations. Other works focus on machine learning methods with more mathematical grounding, e.g., orthogonal projection on high dimensional spaces, but there is no neural correspondence for neuromorphic computing. In this work, we develop a new method with neuronal operations based on lateral connections and Hebbian learning, which can protect knowledge by projecting activity traces of neurons into an orthogonal subspace so that synaptic weight update will not interfere with old tasks. We show that Hebbian and anti-Hebbian learning on recurrent lateral connections can effectively extract the principal subspace of neural activities and enable orthogonal projection. This provides new insights into how neural circuits and Hebbian learning can help continual learning, and also how the concept of orthogonal projection can be realized in neuronal systems. Our method is also flexible to utilize arbitrary training methods based on presynaptic activities/traces. Experiments show that our method consistently solves forgetting for spiking neural networks with nearly zero forgetting under various supervised training methods with different error propagation approaches, and outperforms previous approaches under various settings. Our method can pave a solid path for building continual neuromorphic computing systems.

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Builds on 43 of its 68 references that are in the map. First posted on arXiv 19 Feb 2024.

Builds on

43 of 68 references.

References · 43

  1. 2016Overcoming catastrophic forgetting in neural networkscited by 4,787
  2. 2016Learning without Forgettingcited by 2,496
  3. 2016iCaRL: Incremental Classifier and Representation Learningcited by 2,333
  4. 2017Gradient Episodic Memory for Continual Learningcited by 1,898
  5. 2016Progressive Neural Networkscited by 1,361
  6. 2017Continual Learning with Deep Generative Replaycited by 1,158
  7. 2017Memory Aware Synapses: Learning what (not) to forgetcited by 1,105
  8. 2018Experience Replay for Continual Learningcited by 1,095
  9. 2018Efficient Lifelong Learning with A-GEMcited by 1,058
  10. 2017PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruningcited by 898
  11. 2013An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networkscited by 722
  12. 2017Lifelong Learning with Dynamically Expandable Networkscited by 694
31 more
  1. 2018Overcoming catastrophic forgetting with hard attention to the taskcited by 681
  2. 2017Spatio-Temporal Backpropagation for Training High-performance Spiking Neural Networkscited by 406
  3. 2019Orthogonal Gradient Descent for Continual Learningcited by 369
  4. 2021Gradient Projection Memory for Continual Learningcited by 348
  5. 2021Deep Residual Learning in Spiking Neural Networkscited by 331
  6. 2020Going Deeper With Directly-Trained Larger Spiking Neural Networkscited by 278
  7. 2018SLAYER: Spike Layer Error Reassignment in Timecited by 266
  8. 2016Direct Feedback Alignment Provides Learning in Deep Neural Networkscited by 240
  9. 2016Training Deep Spiking Neural Networks using Backpropagationcited by 205
  10. 2017SuperSpike: Supervised learning in multi-layer spiking neural networkscited by 182
  11. 2018Long short-term memory and learning-to-learn in networks of spiking neuronscited by 181
  12. 2018Continual Learning of Context-dependent Processing in Neural Networkscited by 179
  13. 2021Accurate and efficient time-domain classification with adaptive spiking recurrent neural networkscited by 115
  14. 2022Training High-Performance Low-Latency Spiking Neural Networks by Differentiation on Spike Representationcited by 89
  15. 2022TRGP: Trust Region Gradient Projection for Continual Learningcited by 82
  16. 2016Supervised learning based on temporal coding in spiking neural networkscited by 76
  17. 2022Online Training Through Time for Spiking Neural Networkscited by 49
  18. 2019Temporal-Coded Deep Spiking Neural Network with Easy Training and Robust Performancecited by 46
  19. 2019Temporal Coding in Spiking Neural Networks with Alpha Synaptic Function: Learning with Backpropagationcited by 45
  20. 2021Training Feedback Spiking Neural Networks by Implicit Differentiation on the Equilibrium Statecited by 40
  21. 2018Biologically-plausible learning algorithms can scale to large datasetscited by 37
  22. 2023Towards Memory- and Time-Efficient Backpropagation for Training Spiking Neural Networkscited by 37
  23. 2019Spike-Train Level Backpropagation for Training Deep Recurrent Spiking Neural Networkscited by 32
  24. 2020Online Spatio-Temporal Learning in Deep Neural Networkscited by 32
  25. 2021A Long Short-Term Memory for AI Applications in Spike-based Neuromorphic Hardwarecited by 32
  26. 2021Synaptic metaplasticity in binarized neural networkscited by 29
  27. 2020Brain-inspired global-local learning incorporated with neuromorphic computingcited by 28
  28. 2015A Hebbian/Anti-Hebbian Neural Network for Linear Subspace Learning: A Derivation from Multidimensional Scaling of Streaming Datacited by 26
  29. 2017ASP: Learning to Forget with Adaptive Synaptic Plasticity in Spiking Neural Networkscited by 16
  30. 2023SPIDE: A Purely Spike-based Method for Training Feedback Spiking Neural Networkscited by 8
  31. 2019ODE-Inspired Analysis for the Biological Version of Oja's Rule in Solving Streaming PCAcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 5; 2026: 2.

8 papers.

  1. 2024Continual Learning with Neuromorphic Computing: Foundations, Methods, and Emerging Applicationscited by 8
  2. 2026Beyond representational alignment with brain-guided language models for robust reasoningcited by 4
  3. 2025Replay4NCL: An Efficient Memory Replay-based Methodology for Neuromorphic Continual Learning in Embedded AI Systemscited by 3
  4. 2025Edge Intelligence with Spiking Neural Networkscited by 2
  5. 2025An Attention-based Feature Memory Design for Energy-Efficient Continual Learningcited by 2
  6. 2025Understanding Catastrophic Interference: On the Identifibility of Latent Representationscited by 1
  7. 2026Orthogonal Weight Modification Enhances Learning Scalability and Convergence Efficiency without Gradient Backpropagationcited by 1
  8. 2025Frozen Backpropagation: Relaxing Weight Symmetry in Deep Spiking Neural Networkscited by 0