Skip to content
Code for paper "Successor Uncertainties: Exploration and Uncertainty in Temporal Difference Learning"
Python
Branch: master
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
models initial commit May 22, 2019
.gitignore initial commit May 22, 2019
atari_env.py initial commit May 22, 2019
configs.py initial commit May 22, 2019
losses_functional.py initial commit May 22, 2019
readme.md Update readme.md Dec 3, 2019
replay_buffer.py initial commit May 22, 2019
requirements.txt Security fix -- update version of Pillow Oct 28, 2019
run_atari.py initial commit May 22, 2019
run_test.py initial commit May 22, 2019
utilities.py removed logging command Dec 3, 2019

readme.md

Code for paper "Successor Uncertainties: Exploration and Uncertainty in Temporal Difference Learning" by David Janz*, Jiri Hron*, Przemysław Mazur, Katja Hofmann, José Miguel Hernández-Lobato, Sebastian Tschiatschek. NeurIPS 2019. * Equal contribution

Paper is available at https://arxiv.org/abs/1810.06530.

This code allows for reproduction of the Atari experiments. Click here for code to reproduce the tabular experiments.

To reproduce results, clone && pip install the requirements, then run

python3 run_atari.py --game Enduro

to train a Successor Uncertainties model with parameters as per the paper. This will output training information in tensorboard format to a subdirectory called logs. To obtain test scores, run

python3 /path/to/log_folder output_file.txt

The final score will be output to output_file.txt and progress of testing will be reported to stdout.

You can’t perform that action at this time.