Research Fellow, LLM (School of Computing)
Summary
Research Fellow at NUS to design and evaluate LLM-based agent systems for social simulations that inform policy decisions, using Python, PyTorch, and LLM APIs.
Interested applicants are invited to apply directly at the NUS Career Portal. Please note your application will only be processed if you apply via NUS Career Portal.
NUS Career Portal link: https://careers.nus.edu.sg/job/Research-Fellow%2C-LLM-%28School-of-Computing%29/33924-en_GB/
We regret that only shortlisted candidates will be notified.
Job Description
The National University of Singapore invites applications for the position of Research Assistant in the School of Computing (SoC). We are seeking highly motivated and skilled researchers to drive innovation in persona-based social simulations for aiding policy design.
Position Overview
The research project will focus on building and evaluating novel LLM-based systems for performing social simulations that can aid in policy design. The Research Fellow will work closely with the Principal Investigator and other project members on different aspects of the project including persona creation, agent architecture, and prompt deisgn to the empirical validation of the simulations.
Key Responsibilities
- Develop LLM-based agent architecture representing diverse personas for multilingual and multicultural population
- Build and evaluate persona-based simulation systems for aiding policy design
- Conduct rigorous research and evelop innovative solutions to complex problems
- Publish and present research findings in top academic journals and conferences
- Develop and present demos for your research
- Provide guidance to PhD students and Research Assistants in the group
- Participate in project meetings and contribute to interdisciplinary collaborative efforts
Qualifications
- PhD degree in Computer Science, Artificial Intelligence, or related disciplines
- Prior experience building AI/ML applications, LLM fine-tuning, RAG systems, and agentic frameworks
- High proficiency in programming with Python, PyTorch, LLM APIs and similar frameworks
- Ability to process large-scale datasets, clean data, and apply statistical tools (like Pandas, R or SQL) to evaluate simulation results, e.g., using experiments, is an advantage
- Publication record in venues such as NeurlPS, ICML, ICLR, AAAI, ACL, EMNPL, or comparable conferences and journals
- Good communication skills in English to interact with researchers across the inter-discplinary team
Experience / interest in
- Agent-based modelling or synthetic user research platforms
- Social network principles like preferential attachment, opinion dynamics, and netowrk behaviour to ensure simulated agents interact realistically
Only shortlisted candidates will be notified.