Always dreamed of doing a PhD in artificial intelligence and reinforcement learning? We are seeking a highly motivated student to join our group and work on cutting-edge reinforcement learning research!

This PhD position is fully funded for an initial period of three years through an FWF (Austrian Science Fund) project. The core project aims to tackle a major bottleneck in making reinforcement learning more sample-efficient and stable: learning effectively from different sources of experience (i.e., off-policy data). In particular, we will focus on the semi-gradient bias that affects policy gradient algorithms (Imani, Graves, and White, 2018). You will have the opportunity to advance RL theory, publish at top-tier venues, and ultimately shape your own research agenda. We are looking for independent thinkers who will bring their own taste to the table, propose novel algorithmic directions, and pivot as new discoveries emerge.

Funding Note: Initial funding is for 3 years (standard in Austria). However, we proactively work with our students to secure extensions or bridge funding to ensure a successful, unhurried dissertation.

The project is in collaboration with two leading researchers in reinforcement learning:

• Martha White from the University of Alberta (one of the authors of Off-Policy Actor-Critic, which laid the foundation for SAC, DDPG, and TD3 architectures)
• Carlo d’Eramo from the University of Würzburg (a leading researcher in multitask reinforcement learning and the designer of mushroom-rl).

What we offer:

• Inclusion in the Intelligent Interactive System lab, a large, international research laboratory https://iis.uibk.ac.at/ https://iis.uibk.ac.at/people focused on solving robot learning challenges
• Be part of a wide research network: Having worked in several large international labs (i.e., AIRLab @ Politecnico di Milano, IAS @ TU Darmstadt, and RLAI @ University of Alberta), your advisor can connect you with different research groups. Furthermore, you will join a yearly workshop organized by Prof. Jan Peters and attended by 100+ robot learning researchers from across Europe – a huge networking opportunity!
• You will have access to large-scale clusters for scientific computing (such as LEO5 and IFI-Cluster at the University of Innsbruck, VSC, and Musica) and powerful local machines.
• You can work with robots such as Emika-Franka Panda and Unitree A1, and use motion capture systems (i.e., OptiTrack).
• Last but not least: Innsbruck is an amazing city surrounded by fantastic nature, 30 minutes away from the borders to Germany and Italy, and with an international airport that connects to major European destinations.

Required Skills

• Rigorous background in machine learning, ideally with a focus on RL or numerical optimization
• Fluency in mathematics
• Excellent programming skills
• Analytic problem-solving and communication skills
• Fluency in spoken and written English

Apply Now!

Applications will be reviewed on a rolling basis until the position is filled. For full consideration, please submit your application by the 30th of April 2026.