* Postdoc position in emergent deep net communication *

We are offering one postdoc position for up to 5 years to work on the ALiEN project in a team of 5-to-10 people, including the PI, other postdocs and several PhD students. The selected candidate will receive a competitive salary and enjoy considerable freedom to lead basic research work within the framework of the project.

ALiEN (Autonomous Linguistic Emergence in Neural Networks) is an ERC-funded project investigating spontaneous communication in populations of deep neural networks, with the aim to improve their flexibility in solving complex AI tasks and the interpretability of their behavior. You can find the scientific details in the project proposal, available from the project site (http://marcobaroni.org/alien/).

The 5-year project will start in January 2022 (but post-doc starting date is flexible), and it will be run in the COLT group of Pompeu Fabra University (UPF) in Barcelona (Spain), under the supervision of Marco Baroni. Project funds support a considerable investment in a strong computational infrastructure exclusively dedicated to it.

The COLT group is a research unit of the UPF Linguistics Department specializing in an interdisciplinary approach to computational linguistics. It is currently composed of two permanent faculty members, two postdocs and 6 PhD students. UPF is consistently ranked among the top universities in Spain and as one of the best young universities in the world. It features a rich research ecosystem, including labs focusing on NLP, machine learning and computational evolution. Barcelona is a lively research hub, with top-level company and academic research labs in AI, computer vision and NLP, several located just a short walk from the UPF campus. Barcelona is also famous for good weather, a relaxed cosmopolitan lifestyle, a nice sea-front and a hilly background.

Successful postdocs should have demonstrated the ability to conduct top-tier research in one or more of the following fields: deep learning, natural language processing, computer vision, reinforcement learning. This could be shown, for example, by publications at conferences such as ICLR, ACL or CVPR. They should be curious about deep net behavior and the universal roots of communication. They should enjoy engaging in long-shot basic research, and thrive in an environment that emphasizes friendliness and a good work-life balance over a competitive attitude.

COLT and UPF commit to a safe and inclusive research environment for all of its members. Women and people of marginalized communities are strongly encouraged to apply.

To express interest, please send a CV to mbaroni@gmail.com, including the names and contacts of two references, and briefly explain in the message body why you are interested in this opportunity, and how you see yourself contributing to the team.