This repository provides the implementation of our paper
A Gradient of Complementary Learning Systems Emerges Through Meta‑Learning

It builds upon the official La‑MAML code by Gupta et al. (NeurIPS 2020), extending it to meta‑learn layer‑wise learning rates and representational sparsity in both single‑ and dual‑pathway artificial neural networks for continual learning tasks.

Data Preparation

1. Create a data/ directory in the repository root.
2. Download and place the following datasets into data/ before running experiments:
   • rotated MNIST
   • Fashion MNIST
   • CIFAR‑100

---

Running the Experiments

Example: make scripts executable

chmod +x run_experiments_2layer_models    # for shallow network experiments

---

Citation

If you use this code, please cite both the original La‑MAML paper and our extension:

@article{gupta2020,
  title={La-MAML: Look-ahead Meta Learning for Continual Learning},
  author={Gupta, Gunshi and Yadav, Karmesh and Paull, Liam},
  journal={arXiv preprint arXiv:2007.13904},
  year={2020}
}

@article{zhou2025,
  title={A gradient of complementary learning systems emerges through meta-learning},
  author={Zhou, Zhenglong and Schapiro, Anna},
  journal={bioRxiv},
  year={2025},
  doi={10.1101/2025.07.10.664201}
}

---