Preparatory Exercises for PhD Positions in Safe Agentic/LLM Reasoning via Formal Verification

The University of Manchester; United Kingdom; On-site; GBP 21,000; Part time

Job summary

A leading UK university is offering a 3.5-year PhD studentship focusing on the intersection of formal software verification and Large Language Model safety. Candidates will explore Human-AI interaction and contribute to safety in LLM-powered systems. The position is open to Home (UK) and overseas applicants, with an annual tax-free stipend of £20,780. Applicants should hold at least a 2.1 honours degree or master’s in a relevant discipline, with a strong background in engineering or computer science.

Benefits

Annual tax-free stipend

Tuition fees paid

Qualifications

• Applicants must have relevant science or engineering background.
• Strong understanding of Large Language Model safety is preferred.
• Experience in formal software verification is ideal.

Responsibilities

• Explore novel synergies among Human-AI interaction and optimization.
• Investigate safety and reliability in LLM-powered systems.
• Extend automated reasoning tools to address computational challenges.

Skills

Excellent English proficiency

Expertise in Human-AI interaction

Knowledge of machine learning

Formal software verification

Education

Minimum 2.1 honours degree or master’s in relevant discipline

Job description

Organisation/Company The University of Manchester Department Computer Science Research Field Computer science Researcher Profile First Stage Researcher (R1) Positions PhD Positions Application Deadline 6 Feb 2026 - 23:59 (Europe/London) Country United Kingdom Type of Contract Temporary Job Status Other Job Status Extra Information 3.5-year PhD studentship Offer Starting Date 1 Apr 2026 Is the job funded through the EU Research Framework Programme? Horizon Europe (other) Reference Number 101226927 Is the Job related to staff position within a Research Infrastructure? No

Offer Description

The target is to explore novel synergies among natural language-based Human-AI interaction, nature-inspired optimization and exploration, the social impacts of Human-AI cooperation on team dynamics, and advanced machine learning in the engineering domain. Generative AI substantially changes the way engineers interact with and benefit from AI and access domain-specific knowledge, marking a phase transition that not only enhances the productivity of engineering teams but also expands the complexity of solvable design problems.

We have two exciting PhD positions at the intersection of formal software verification and Large Language Model (LLM) safety, focusing on extending state-of-the-art logic-based automated reasoning tools such as ESBMC (https://github.com/esbmc/esbmc ) to address safety and reliability challenges in agentic reasoning systems. Successful candidates will investigate novel approaches that use abstract interpretation, model checking, constraint programming, and fuzzing techniques to ensure safety and reliability in LLM-powered agentic systems.

Applicants should have, or expect to achieve, at least a 2.1 honours degree or a master’s (or international equivalent) in a relevant science or engineering related discipline.

Languages ENGLISH Level Excellent

Research Field Computer science Years of Research Experience 4 - 10

Additional Information

This 3.5-year PhD studentship is open to Home (UK) and overseas applicants. The successful candidate will receive an annual tax-free stipend set at the UKRI rate (£20,780 for 2025/26; subject to annual uplift), and tuition fees will be paid. We expect the stipend to increase each year.The start date is April 2026.

Eligibility criteria

Applicants should have, or expect to achieve, at least a 2.1 honours degree or a master’s (or international equivalent) in a relevant science or engineering related discipline.

Applicants should follow the Mobility Rule: DCs have not been resident in the country of the recruiting beneficiary for more than 12 months in the last 36 months.