Named in honour of Alan Turing, the Institute is a place for inspiring, exciting work and we need passionate, sharp, and innovative people who want to use their skills to contribute to our mission to make great leaps in data science and AI research to change the world for the better.

Please find more information about us here.

In 2022, the Alan Turing Institute signalled its intention to establish a portfolio of foundational AI research, which would complement the strengths of the institute around applications of AI and AI policy. An initial portfolio of research in foundation models, game theory, and probabilistic programming will be launched in early 2023. Each of these areas is called a ‘Pillar’. It is intended that this portfolio will complement the UK’s current activity, rather than duplicating existing efforts, and aiming to promote emerging new areas that show promise for the future.

FOUNDATION MODELS

Foundation models are large ML models trained on large, broad data sets. Foundation models such as GPT have been shown to have remarkable capabilities for generating realistic natural language, and, to some extent, capabilities for problem solving and common-sense reasoning. Developing a Turing Foundation Model is beyond our present capacity. Instead, we therefore propose work aimed at developing Turing expertise around the problem of precisely understanding the capabilities of such models. The main issue we aim to address is that of *benchmarking* such models: although such models appear to be very capable in some respects, they fail on apparently simple tasks, in unpredictable ways. In short, we don't have a clear understanding of the capabilities and shortcomings of such systems - which raises concerns for their use.

In this activity, we will work with Foundation Models such as GPT focussing on three key questions:

• Learned values. In this work we will investigate the extent to which such models learn human values. To what extent can it be said to understand the human values raised in such scenarios? Is it consistent in the application of such values? Does it hold such values immutably, or are they malleable?
• Common sense reasoning. Existing Foundation Models appear to have some common-sense reasoning capability, but to what extent do they understand (e.g.) naive physics, spatial and temporal reasoning, the concept of agency, and so on?
• Theory of mind. Much human reasoning is social, involving the beliefs and aspirations of others. To what extent can this capability be acquired by training on textual data sets?

The Turing are hiring a team of Research Associates to support and enable the delivery of the main research themes under the direction of the Turing Programme Director Professor Michael Wooldridge, and the project PIs. The team will collectively have a broad range of expertise, and we will be hiring based on varied skills required for each specific challenge.

We are looking for a Senior Research Associate to support and enable the delivery of the Foundation Model theme, under the direction of Anthony Cohn, and in collaboration with Michael Wooldridge and Nigel Shadbolt.

ROLE PURPOSE

The successful candidate will primarily focus on evaluating the extent to which, and the conditions under which existing Foundational Models can support common-sense reasoning (such as naive physics, spatial and temporal reasoning, the concept of agency, and causality). This work will be taken in parallel and in collaboration with researchers working on other aspects of the Foundation Models Pillar.

The candidate will join a vibrant team of researchers and will have opportunities to engage with cutting-edge projects and experts at leading universities.

DUTIES AND AREAS OF RESPONSIBILITY

• Develop and implement new methods for representing, transforming, and analysing foundation models.
• Develop new project concepts and proposals and assume overall leadership for the delivery of research projects from inception to completion.
• Run computational experiments to analyse common sense reasoning and understand the efficacy of new and existing analysis methods.
• Define the research direction in collaboration with Project Investigators
• Keep on top of the state of the art in the relevant literatures, in particular at the intersection of Foundation Models and AI in the area of common sense reasoning.
• Take the lead on writing up findings as they emerge, producing reports, and developing publications in peer reviewed journals, in collaboration with the research team.
• Help to write academic research papers and disseminate our research output to the research community, e.g., by giving talks at international conferences in computing science and other relevant disciplines, and co-authoring articles for journals.
• Present, disseminate and explain our work at internal and external events hosted by the Alan Turing Institute.

• A PhD or equivalent qualification/experience in Mathematics, Computer Science or a closely related discipline.
• A solid background in two or more of the following: use of Foundational Models such as GPT; AI-based methods for common sense reasoning methods such as naive physics, spatial and temporal reasoning, and causality; evaluation of AI systems.
• Track record of the ability to initiate, develop and deliver high quality research aligned with the research strategy and any industrial stakeholders and to publish in peer reviewed journals and conferences.
• Hands-on experience with Machine Learning methods
• Ability to communicate complex, specialist or conceptual information clearly and persuasively to diverse audiences.
• Ability to organise working time, take the initiative, and carry out research independently.
• Participates in networks within the organisation or externally to share knowledge and information in order develop practice or help others learn.
• Ability to work effectively across disciplinary boundaries, both as part of an interdisciplinary team and in close collaboration with external partners in different disciplines.

APPLICATION PROCEDURE

If you are interested in this opportunity, please click the apply button below. You will need to register on the applicant portal and complete the application form including your CV and covering letter. If you have questions about the role or would like to apply using a different format, please contact us at [email protected].

If you are applying for more than one role at the Turing, please note that only one Cover Letter can be visible on your profile at one time. If you wish to apply for multiple roles and do not want to overwrite your existing Cover Letter, please apply for the role using the button below and forward your additional cover letter directly to [email protected] quoting the job title.

If you are an internal applicant and wish to apply, please send your CV and Cover Letter directly to [email protected] and your application will be considered.

CLOSING DATE FOR APPLICATIONS: Sunday 12 November 2023 at 23:59 (London, UK BST)

We reserve the right to close this vacancy early or to interview suitable candidates before the closing date.

TERMS AND CONDITIONS

This full-time post is offered on a 2-year fixed-term basis. The annual salary is £53,576 - £55,125 plus excellent benefits, including flexible working and family friendly policies, https://www.turing.ac.uk/work-turing/why-work-turing/employee-benefits

The post can be based either at The Alan Turing Institute site in London, or at the University of Leeds. In either case you will need to travel to the other site when required (travel expenses will be paid as appropriate).

The Alan Turing Institute is based at the British Library, in the heart of London’s Knowledge Quarter. We expect staff to come to our office at least 4 days per month. Some roles may require more days in the office; the hiring manager will be able to confirm this during the interview.

EQUALITY, DIVERSITY, AND INCLUSION

The Alan Turing Institute is committed to creating an environment where diversity is valued and everyone is treated fairly. In accordance with the Equality Act, we welcome applications from anyone who meets the specific criteria of the post regardless of age, disability, ethnicity, gender reassignment, marital or civil partnership status, pregnancy and maternity, religion or belief, sex and sexual orientation.

We are committed to building a diverse community and would like our leadership team to reflect this. We therefore welcome applications from the broadest spectrum of backgrounds.

We are committed to making sure our recruitment process is accessible and inclusive. This includes making reasonable adjustments for candidates who have a disability or long-term condition. Please contact us at [email protected] to find out how we can assist you.

Please note all offers of employment are subject to obtaining and retaining the right to work in the UK and satisfactory pre-employment security screening which includes a DBS Check. Full details on the pre-employment screening process can be requested from [email protected].