Papers of the day   All papers

Behaviour Suite for Reinforcement Learning

Comments

Aug 13 2019 DeepMind

We are excited to release Behaviour Suite for Reinforcement Learning, or ‘bsuite’ – a collection of carefully-designed experiments that investigate core capabilities of RL agents GitHub: http://github.com/deepmind/bsuite Paper: https://arxiv.org/abs/1908.03568v1
2 replies, 1313 likes


Aug 13 2019 Ian Osband

Really excited to release #bsuite to the public! - Clear, scalable experiments that test core #RL capabilities. - Works with OpenAI gym, Dopamine. - Detailed colab analysis - Automated LaTeX appendix Example report: http://bit.ly/bsuite-agents https://t.co/IHuu5Gjmmk
0 replies, 182 likes


Aug 13 2019 koray kavukcuoglu

Open sourcing bsuite. Automated evaluation and analysis of agents on RL benchmarks. We hope this will help reproducible and accessible research on core problems in RL. Looking forward to seeing results from the community! https://t.co/Ixz1xXbGh3
0 replies, 63 likes


Aug 13 2019 Danilo J. Rezende

It is not trivial to make proper and fair experiments in RL. This work from @IanOsband et al is a great step in that direction.
1 replies, 53 likes


Aug 13 2019 Dumitru Erhan

Love this stuff! Performance of agents cannot (and should not!) be reduced to just "sample efficiency" or "final performance", a fine-grained understanding of the various trade-offs is almost always a good thing.
1 replies, 21 likes


Aug 13 2019 Max Jaderberg

Check out bsuite: open-source way to analyse the capabilities of RL algorithms. Baselines (including OpenAI and Dopamine) with full training script, even launching experiments on GCP and fancy latex appendixes for your papers. Super cool from @IanOsband et al
1 replies, 21 likes


Aug 13 2019 Jane Wang

Really cool resource put together by @IanOsband and colleagues for characterizing core capabilities of RL agents! Try it out here: http://bit.ly/bsuite-tutorial-colab
0 replies, 12 likes


Aug 20 2019 Isaac Kamlish

Interesting approach to fairer and more interpretable experimentation within RL. Well done @IanOsband !
0 replies, 4 likes


Aug 14 2019 Anugraha Sinha

@__MLT__ @suzatweet @yoovraj
0 replies, 1 likes


Content