Geek.AI Group

UK-China joint research group for the next generation of AI

Email: j.wang [AT] cs.ucl.ac.uk

Artificial Intelligence (AI) techniques have been existing in various scenarios of our life and improving the life quality and work efficiency. Generally, there are two categories of AIs, namely the AIs making prediction (or recognition) of the data, and the AIs making decisions (or controls) via interacting with the environment. With the recent success of superhuman AIs like AlphaGO in the game of Go and DeepStack in poker games, more and more attention is now focused on AI for decision making, which heavily involves deep reinforcement learning techniques. Furthermore, regarding the real world as of multiple agents interacting with each other, multi-agent (deep) reinforcement learning starts to earn more attention. One can believe that the Artificial Collective Intelligence (ACI) revealed from multi-agent systems powered by deep reinforcement learning and game theory is the pursuit and the central topic of the next generation of AI.

Geek.AI is a UK-China joint research team dedicated on fundamental research topics of multi-agent reinforcement learning and its potential for real-world applications. On-going research topics include game AI coordination, learning to communicate, RL with massive number of agents, self-play, learning to design multi-agent environment, personality learning for AI agents, GANs for discrete data generation etc.

Featured Open Source Projects


link
MAgent: A Many-Agent Reinforcement Learning Platform for Artificial Collective Intelligence
Lianmin Zheng, Jiacheng Yang, Han Cai, Weinan Zhang, Jun Wang, Yong Yu
Demos in NIPS 2017 & AAAI 2018.  

link
Texygen: A Benchmarking Platform for Text Generation Models
Yaoming Zhu, Sidi Lu, Lei Zheng, Jiaxian Guo, Weinan Zhang, Jun Wang, Yong Yu
In progress.

Publications


pdf
Efficient Architecture Search by Network Transformation
Han Cai, Tianyao Chen, Weinan Zhang, Yong Yu, Jun Wang
AAAI 2018.  

pdf
A Neural Stochastic Volatility Model
Rui Luo, Weinan Zhang, Xiaojun Xu, Jun Wang
AAAI 2018.  

pdf
Long Text Generation via Adversarial Training with Leaked Information
Jiaxian Guo, Sidi Lu, Han Cai, Weinan Zhang, Yong Yu, Jun Wang
AAAI 2018.

pdf
An Empirical Study of AI Population Dynamics with Million-agent Reinforcement Learning
Yaodong Yang, Lantao Yu, Yiwei Bai, Jun Wang, Weinan Zhang, Ying Wen, Yong Yu
NIPS 2017 Aligned AI Workshop.

pdf
IRGAN: A Minimax Game for Unifying Generative and Discriminative Information Retrieval Models
Jun Wang, Lantao Yu, Weinan Zhang, Yu Gong, Yinghui Xu, Benyou Wang, Peng Zhang, Dell Zhang
SIGIR 2017.   (Best Paper Honorable Mention Award)

link
SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient
Lantao Yu, Weinan Zhang, Jun Wang, Yong Yu.
AAAI 2017.

Researchers


link
Jun Wang
Professor, University College London, UK

link
Dell Zhang
Senior Lecturer, Birkbeck, University of London, UK

link
Weinan Zhang
Assistant Professor, Shanghai Jiao Tong University, China

link
Kan Ren
Ph.D., Shanghai Jiao Tong University, China

News


7 Feb. 2018
Texygen platform is published on Github.

8 Dec. 2017
MAgent platform demonstrated in NIPS 2017.

02 Mar. 2017
Group website launched.