Now recruiting: Paid 6-month Research Assistant · LLM Agents & Information RetrievalView position ↗
Zhongtian Sun
Artificial intelligence · Machine learning · Research

Zhongtian Sun

Trustworthy AI, representation learning and structured reasoning.

I develop machine-learning methods to understand and improve how AI systems learn, reason and generalise, with applications across healthcare, finance and education.

Lecturer / Assistant Professor in AI & Machine Learning · University of Kent ↗

Visiting Fellow · University of Cambridge ↗

Mila AI Policy Fellow

Work with me ↗Explore publications ↗

Research Master’s and PhD students, visiting researchers and research interns are welcome to get in touch. I also welcome academic and industry collaborations — including ideas beyond my current research themes.

LinkedIn ↗
Portrait of Dr Zhongtian Sun
Opportunities

Work with me

I welcome motivated researchers at different career stages, including students with their own research questions. My current research interests offer a starting point, not a fixed list of projects.

Now recruiting · Paid · 6 monthsResearch Assistant
LLM Agents & Information Retrieval

I am recruiting for a paid, six-month Research Assistant position on an applied AI project with an external healthcare-sector organisation. The role involves developing, prototyping and evaluating LLM-based AI systems for real-world applications.

Relevant experience: Python, large language models, AI agents, information retrieval or retrieval-augmented generation (RAG). Experience with LLM evaluation, software development or applied NLP is an advantage.

To apply, email a CV, a brief summary of relevant experience and links to previous projects or code (if available). Apply by email ↗

Research Master’s

MSc by Research applicants, research-oriented Master’s students and dissertation projects in AI and machine learning. I welcome original project ideas, including those extending beyond my current topics. Research degrees at Kent ↗

PhD students

Doctoral research in trustworthy AI, geometric and structured learning, and related areas. Students are encouraged to propose their own research questions and approaches. Cross-institutional research and co-supervision are also part of my work.

Interns & visiting researchers

Short-term research projects, research internships and visiting-student collaborations, including researchers based at other institutions. Projects are shaped around shared interests and appropriate supervision.

Industry-linked research

Students and researchers interested in research with external partners are welcome to make contact. I also work with companies on AI-related research and interdisciplinary projects.

Postdocs & academic partners

Joint publications, research projects, interdisciplinary partnerships and fellowship collaborations with postdoctoral researchers and academic colleagues.

Specific project participation, visits, funding and formal supervision arrangements are agreed individually with the relevant partners and institutions.

Research

Research interests

My work spans trustworthy AI, geometric and topological learning, structured reasoning and causal machine learning.

Trustworthy, reliable & agentic AI

AI reliability, safety, interpretability, uncertainty, autonomous agents and multi-agent systems.

Geometric & topological learning

Neural representations, geometric and topological machine learning, graphs, hypergraphs and sheaf-based learning.

Structured & knowledge-grounded AI

Neuro-symbolic AI, structured reasoning, knowledge graphs, language models and world models.

Causal & applied machine learning

Causal methods, robust generalisation and applications in financial systems, healthcare, education and scientific AI.

Publications

Selected publications

Selected work across trustworthy AI, geometric learning, structured models and applied machine learning.

2026

A Topological Signature of Memorization under Noisy Labels

Z. Sun, F. Mo, P. Tiwari, K. Xia

NeurIPS 2026 · Main conference · Accepted

Full publication list on my Kent profile ↗

Collaboration

Research & collaboration

Research progresses through shared expertise, open exchange and work across disciplinary boundaries.

I collaborate with researchers across universities, research institutes and industry. This work includes joint publications, doctoral research and interdisciplinary projects, connecting fundamental machine learning with applications in healthcare, finance and beyond.

At Cambridge, I work closely with Professor Pietro Liò. Our collaboration includes published research, co-supervision of doctoral students, and ongoing work in representation learning and agentic AI. I also collaborate with doctoral researchers at Oxford on joint publications, research proposals and projects.

Examples of joint work with Pietro Liò include WWW 2025 ↗ and MONEY · AI Open 2023 ↗. Other publications are listed above.

Interested in collaborating on research or an industry project?Get in touch ↗
About

Background & service

My experience spans academic machine learning, interdisciplinary research and financial applications.

About me

I received my PhD in Computer Science from Durham University, following an MSc in Business Analytics at Warwick and a BSc in Finance, Accounting and Management at Nottingham.

Before joining Kent, I conducted research at Cambridge on machine learning for neuroscience and at Oxford on large language models, knowledge graphs and reasoning. I have also worked in asset management and served as CTO at an AI-in-finance startup.

At Kent I teach and supervise across machine learning, deep learning, natural language processing and software development.

Kent ↗Cambridge ↗ORCID ↗ LinkedIn ↗

Selected academic service

Selected talks