Postdoctoral Researcher in Implicit Neural Networks, Lyapunov Stability, and Data-Driven Control Inria Saclay in United Kingdom

Postdoctoral researcher opportunity in implicit neural networks, Lyapunov stability, data-driven control, and machine learning at the interface of artificial intelligence and control systems.

The position is a collaboration between Inria Saclay DISCO (CentraleSupélec / Paris-Saclay ecosystem) and University College London through the UCL-Inria Joint Centre. The research focuses on whether implicit neural networks can provide stronger stability and safety certificates than standard feedforward architectures, with particular attention to the expressivity of neural parameterizations of Lyapunov functions, smooth vs. non-smooth activations, neural network verification, and training strategies for implicit models.

The postdoctoral researcher will spend 12 months in Palaiseau, France, and 12 months in London, UK. The role is supervised by Prof. Giorgio Valmorbida and Prof. Akin Delibasi. The project includes developing theoretical and computational methods for control Lyapunov functions, optimization-based control design, and comparing implicit and feedforward neural networks under equivalent parameterization constraints.

Eligibility highlights: PhD or equivalent in control theory, artificial intelligence, or a closely related field; strong background in nonlinear control, Lyapunov analysis, and optimization; advanced programming in Python and MATLAB; fluent English; French is an asset; strong publication record expected.

Funding and conditions: fixed-term 2-year contract with a gross salary of €2,788/month. Benefits include subsidized meals, transport reimbursement, leave allowances, teleworking options, equipment, training, and social security coverage.

Deadline: 2026-10-01. Start date: 2026-11-02.

Apply via the official Inria job portal. The post also provides a research proposal download link and notes that applications submitted through other channels may not be processed.