SNAP: Sigmoidal-Neuronal-Adaptive-Plasticity

SNAP is an approximation to Long-Term Potentiation for Artificial Neural Networks to reduce catastrophic forgetting.

Read the paper»

Getting Started

Prerequisites

The project uses uv to manage and lock project dependencies for a consistent and reproducible environment. If you do not have uv installed on your system, visit this page for installation instructions.

Note: If you have pip you can just invoke:

pip install uv

Installation

# Clone the repo
git clone git@github.com:sebastian9991/SNAP-Sigmoidal-Neuronal-Adaptive-Plasticity.git

# Enter the repo directory
cd SNAP-Sigmoidal-Neuronal-Adaptive-Plasticity

# Install core dependencies into an isolated environment
uv sync

Usage

Running Full SNAP Experiments

./run_all_experiments

License

Distributed under the MIT License. See LICENSE.txt for more information.

Citation

@article{xu2024snapstoppingcatastrophicforgetting,
      title={SNAP: Stopping Catastrophic Forgetting in Hebbian Learning with Sigmoidal Neuronal Adaptive Plasticity}, 
      author={Tianyi Xu and Patrick Zheng and Shiyan Liu and Sicheng Lyu and Isabeau Prémont-Schwarz},
      year={2024},
      eprint={2410.15318},
      archivePrefix={arXiv},
      primaryClass={cs.NE},
      url={https://arxiv.org/abs/2410.15318}, 
}