Continuous Machine Learning

We are looking for a researcher in Artificial Intelligence who will integrate knowledge of current deep learning with what is known about continuous incremental learning in cognitive systems. You should have affinity with non-standard neuron models (e.g., spiking neurons). Rather than training single-trick models, the challenge is to develop systems that can build upon existing skills (e.g., Hinton's capsules) and adapt to new tasks in a compositional manner. New paradigms for performance evaluation and benchmarking need to be designed, with a possible translation to neuromorphic hardware platforms in mind. The position is embedded in the Autonomous Perceptive Systems group of the Bernoulli Institute, with strong collaborations with materials science, cognitive neuroscience and computer science.

For more information about the position click here and to apply click here.