Memory-Augmented Network System

A Memory-Augmented Network System is a ANN System that implements a memory network algorithm.

• Example(s):
  • Memory Networks (Torch/Matlab): https://github.com/facebook/MemNN
  • python-babi: MemN2N implemenation on bAbI tasks with very nice interactive demo. https://github.com/vinhkhuc/MemN2N-babi-python
  • theano-babi: MemN2N implementation in Theano for bAbI tasks. https://github.com/npow/MemN2N
  • tf-lang: MemN2N language model implementation in TensorFlow. https://github.com/carpedm20/MemN2N-tensorflow
  • tf-babi: Another MemN2N implementation of MemN2N in TensorFlow, but for bAbI tasks. https://github.com/domluna/memn2n
  • Simulation tasks generator: https://github.com/facebook/bAbI-tasks
• See: ConvNet System.

---

---

References

2017

• https://github.com/vinhkhuc/MemN2N-babi-python
  • QUOTE: This project contains implementations of memory augmented neural networks. ...

---