Research Associate in Safe Reinforcement Learning

Job description

About the role

We are looking for motivated individuals to Join the Formal Methods in AI (FMAI) lab at Imperial College London, led by Dr. Francesco Belardinelli , in a fully funded postdoctoral research role to lead transformative research in formal methods for safe reinforcement learning.

Overview The FMAI lab at Imperial is seeking highly motivated and talented Postdoctoral Research Associates (PDRAs/PostDocs), who have demonstrated competence in conducting cutting-edge research. The position is fully funded in the context of Dr. Francesco Belardinelli's ARIA project Enforcing Safety in Cyber-Physical Systems via Proof Certificates , and focus on the design, development, and application of Safe RL algorithms as well as their verification via Proof Certificates , including monitoring and shielding of cyber-physical systems.

AI-powered cyber-physical systems must operate continuously and reactively in safety-critical environments. Failures pose severe economic risks, even cost human lives.

In recent years, Safe RL has been developed to apply RL techniques in safety-critical environments. However, current methods primarily provide finite-horizon, statistical, or asymptotic guaranties, and fail to ensure strict safety compliance at runtime. This creates a fundamental gap between scalable learning and certifiable safety.

To address this gap, this project aims at developing Certified Reinforcement Learning , a neuro-symbolic framework for learning safe controllers in real-world cyber-physical systems that leverages the scalability and adaptability of RL, while providing the formal, verifiable guaranties associated with Formal Methods.

The proposed methodology will be implemented in the MASA-Safe-RL library- an open-source platform for Safe RL currently being developed at the FMAI lab.

What you would be doing

Within the project, y ou will conduct original research in the new and exciting field of Formal Methods for Safe RL and explore its applications across cyber-physical systems . You will develop novel algorithms that leverage proof certificates . In doing so, you will collaborate with a team of expert researchers in reinforcement learning, formal methods , strategy synthesis, multi-agent systems , and related fields. We strive in publishing in top-tier conferences and journals.

What we are looking for

• " data-aria-posinset="1" data-aria-level="1"> Self-driven and motivated individuals with genuine love for at least one of Formal Methods/Reinforcement Learning, possibly both, with a drive to learn about the other area.

• The applicant is also expected to have a strong track record in top conferences and journals in the field of Formal Methods/Reinforcement Learning, such as AAAI, AAMAS, IJCAI, NeurIPS, ICML, ICLR etc.

• We expect excellent skills in mathematics, especially knowledge in formal methods, stochastic systems and processes, the foundations of deep learning.
• Experience coding with deep learning libraries such as Pytorch/JAX is essential.
• Applicants must hold a PhD in computer science, mathematics or equivalent experience.

Please see job description for a full list of requirements.
"},{

What we can offer you

• " data-aria-posinset="1" data-aria-level="1"> The opportunity to continue your career at a world-leading institution and be part of our mission to continue science for humanity.

• Grow your career: gain access to Imperial's sector-leading dedicated career support for researchers as well as opportunities for promotion and progression.

• Sector-leading salary and remuneration package (including 41 days off a year and generous pension schemes).

• Be part of a diverse, inclusive and collaborative work culture with various staff networks and resources to support your personal and professional wellbeing.

"},{

Further information

Full-time, Fixed term contract to start as soon as possible up to 31 st May 2028.

*Candidates who have not yet been officially awarded their PhD will be appointed as Research Assistant within the salary range £45,399 - £48,876 per annum.

In addition to completing the online application candidates should attach:

• " data-aria-posinset="1" data-aria-level="1"> A full CV with a list of all publications

• A 1-page research statement i ndicating what you see are interesting research issues relating to the above post and why your expertise is relevant.

Informal enquiries related to the position should be directed to Dr Francesco Belardinelli [email protected]

For queries regarding the application process contact Jamie Perrins: [email protected]
"}