PhD on Reactive Bayesian AI Agents

In this PhD project you will develop reactive, Bayesian AI agents that run on portable devices. We take inspiration from how the brain spontaneously reacts to sensory input, as formally laid out by the 'Free Energy Principle' (https://en.wikipedia.org/wiki/Free_energy_principle(FEP). We develop Bayesian AI agents that, similar to the brain, learn purposeful behavior solely through environdmental interactions.To support this research, we are developping a custom reactive probabilistic Programming toolbox named Rxlnfer (see http://rxinfer.ml). RxInfer supports real-time reactive Bayesian inference for AI agents. In your PhD research you will advance both fundamental and practical development of RxInfer and seek novel application areas for reactive probabilistic programming.

Job Description

This PhD project is funded by the ROBUST program (https://icai.ai/labs-robust/) that aims to develop trustworthy AI tools for today's big societal challenges. One of these challenges concerns improving the participation of hearing-impaired persons in challenging work and social settings. In this PhD project, you will develop probabilistic programming tools for real-time AI agents that support situated (i.e., real-time, in-situ) development of personalized audio processing algorithms for hearing aid clients. Your algorithms will be implemented on portable devices and operate under computational and energy-consumption constraints.

An substantial part of the PhD research will be devoted to further development of RxInfer (http://rxinfer.ml), which is a high-quality toolbox-under-development for automating real-time Bayesian inference. Your work will partly consist of developing and coding fundamental (Bayesian) AI tools, and partly on applying these tools to developing AI agents that learn to personalize (recommend) hearing aid algorithms through situated interactions between the agent and a human client.  Therefore, for a perfect fit with this position, you should have a keen interest and background in both quality software development and in Bayesian machine learning methods. RxInfer is based on a Reactive Programming framework and coded in Julia, see http://rxinfer.ml.

You will work in the BIASlab team (http://biaslab.org) in the Electrical Engineering department at TU/e. This lab focuses its research activities on transferring a leading physics-based theory about computation in the brain, the Free Energy Principle (FEP), to practical use in synthetic AI agents. During this project you will closely collaborate with other BIASlab researchers, as well as with project team members at the Human Technology Interaction lab (https://tinyurl.com/2jno83f6), and with our industrial hearing device partner GN Hearing.

Key areas of interest include software development, Bayesian machine learning, probabilistic graphical models (factor graphs), ,signal processing and computational neurosciences.

This research project requires a multidisciplinary approach and draws from Bayesian machine learning, computational neuroscience, and professional-level software development. See this youtube presentation (https://youtu.be/QYbcm6G_wsk) on Natural Artificial Intelligence for more information about our research.

A substantial part of the PhD research will be devoted to further development of RxInfer (http://rxinfer.ml), which is a high-quality toolbox-under-development for automating real-time Bayesian inference. Your work will partly consist of developing and coding fundamental (Bayesian) AI tools, and partly on applying these tools to developing practical AI agents.

RxInfer is based on a Reactive Programming framework and coded in Julia. For a perfect fit with this position, you should have a keen interest and background in both quality software development and in Bayesian machine learning methods.

This PhD project is funded by the ROBUST program (https://icai.ai/labs-robust/) that aims to develop trustworthy AI tools for today's big societal challenges. A particular application area concerns agents that support situated (i.e., real-time, in-situ), personalized audio processing algorithms for hearing aid clients. Your algorithms will be implemented on portable devices and operate under computational and energy-consumption constraints.

You will work in the BIASlab team (http://biaslab.org) in the Electrical Engineering department at TU/e. This lab focuses its research activities on transferring a leading physics-based theory about computation in the brain, the Free Energy Principle (FEP), to practical use in synthetic AI agents. During this project you will closely collaborate with other BIASlab researchers, as well as with project team members at the Human Technology Interaction lab (https://tinyurl.com/2jno83f6), and with our industrial hearing device partner GN Hearing.

A meaningful job in a dynamic and ambitious university, in an interdisciplinary setting and within an international network. You will work on a beautiful, green campus within walking distance of the central train station. In addition, we offer you:

• Full-time employment for four years, with an intermediate evaluation (go/no-go) after nine months. You will spend 10% of your employment on teaching tasks.
• Salary and benefits (such as a pension scheme, paid pregnancy and maternity leave, partially paid parental leave) in accordance with the Collective Labour Agreement for Dutch Universities, scale P (min. €2,770 max. €3,539).
• A year-end bonus of 8.3% and annual vacation pay of 8%.
• High-quality training programs and other support to grow into a self-aware, autonomous scientific researcher. At TU/e we challenge you to take charge of your own learning process.
• An excellent technical infrastructure, on-campus children's day care and sports facilities.
• An allowance for commuting, working from home and internet costs.
• A Staff Immigration Team and a tax compensation scheme (the 30% facility) for international candidates.