/rtrl-elstm

Official repository for the paper "Exploring the Promise and Limits of Real-Time Recurrent Learning" (ICLR 2024)

Primary LanguagePython

Real-Time Recurrent Learning with eLSTM

This is the official repository containing code for the paper:

Exploring the Promise and Limits of Real-Time Recurrent Learning (ICLR 2024)

Contents

  • diagnostic directory contains code for the copy task (Sec 4.1)
  • reinforcement_learning directory contains code for the RL experiments (Sec. 4.2 and 4.3)

Please refer to the readme file in each directory for further instructions. Separate license files can be found in each directory.

BibTex

@inproceedings{irie2023exploring,
  title={Exploring the Promise and Limits of Real-Time Recurrent Learning},
  author={Irie, Kazuki and Gopalakrishnan, Anand and Schmidhuber, J{\"u}rgen},
  booktitle={International Conference on Learning Representations (ICLR)},
  address={Vienna, Austria},
  month=may,
  year=2024
}