University of Liverpool -- Postdoc Position

The department of Computer Science at the University of Liverpool is looking for a motivated and enthusiastic individual to work on an EPSRC project "EnnCore: End-to-End Conceptual Guarding of Neural Architectures" with Dr Xiaowei Huang ( Founded in 1881 as the original `redbrick’, the University of Liverpool is one of the UK’s leading research institutions with an annual turnover of £480 million, including £133 million for research. Liverpool is ranked in the top 1% of higher education institutions worldwide and is a founding member of the prestigious Russell Group, comprising the leading research universities in the UK. The 2014 Research Excellence Framework -- the system for assessing the quality of research in UK higher education institutions -- rated 97% of the Department of Computer Science’s research as being world-leading or internationally excellent, the highest proportion of any Computer Science department in the UK.

You will enjoy a vibrant environment at Liverpool, where in the Department of Computer Science we have an active group of 8-10 researchers working on the intersection of Formal Methods, Machine Learning, and Robotics, aiming to provide certification, assurance, and interpretability to machine/deep learning enabled systems.

The group is currently running other active projects such as EU H2020 project on "FOCETA - Foundations for Continuous Engineering of Trustworthy Autonomy" and Dstl Project on "Test Metrics for Artificial Intelligence". There are plenty of opportunities for you to engage with other projects and other partners. This particular position will include collaboration with our collaborators at Manchester.

You will enjoy developing theories, algorithms, and tools for the verification of deep neural networks and/or the interpretability (explainable AI) of deep neural networks. This might include techniques based on SMT solvers, abstract interpretation, global optimisation methods, neural-symbolic approaches, etc., or techniques that can provide better robustness and generalisation ability to deep learning. In addition to the verification and interpretation of neural network models, you might also consider the verification and analysis of the underlying source code implementation of the neural network.

The group is running an Autonomous Cyber Physical Systems lab, with dedicated access to experimental environments and GPU servers. Therefore, we can accommodate the needs of practical experiments and demonstrations.

You should have, or be about to obtain, a PhD in Computer Science or a closely related field together with an excellent track record of international publications in either the foundation of verification and validation of neural networks or the theoretical analysis of deep learning. Examples of fields of interests are:

To formally apply, please follow the instruction provided in the link:

Please feel free to contact me for more information and informal feedback of your application.