EPSRC iCASE Studentship: Embark on a fully-funded PhD journey with an EPSRC iCASE Studentship focused on advancing the assurance of Large Language Models (LLMs) through the infusion of knowledge graphs. This prestigious 4-year opportunity is funded by QinetiQ, a subsidiary of Naimuri, and the Engineering and Physical Sciences Research Council (EPSRC). The project aims to enhance LLM performance, trustworthiness, and factuality using Neuro-Symbolic AI, a cutting-edge approach combining neural network-based technologies with symbolic systems like Knowledge Graphs.
- 📅 Application Deadline: March 20, 2024
- 🤖 Project Overview:
- QinetiQ is developing an AI Assurance Framework for responsibly utilizing Large Language Models (LLMs) in UK Defence and government agencies.
- Aims to assure LLM products and enhance performance using Neuro-Symbolic AI.
- Focus on combining neural network-based technologies with symbolic systems like Knowledge Graphs.
- Five stages to explore using knowledge graphs to detect hallucinations, falsehoods, contradictions, and knowledge gaps in LLM outputs.
- Enhancements in LLM inference, training, and architecture to improve model quality.
- 🎓 PhD Program Details:
- PhD in Computer Science within the Responsible Artificial Intelligence Group at the University of Hull.
- Prestigious EPSRC iCASE studentship co-funded by Naimuri (QinetiQ subsidiary).
- Active participation in the program, time management, and organization skills required.
- Involvement in research proposal, literature reviews, studies, and thesis writing.
- Benefits from the University of Hull’s Postgraduate Training Scheme.
- 📝 How to Apply:
- Personal statement required, addressing motivation, interest in the project, matching skillset, and future research directions.
- Full-time study, starting in May 2024.
- 💰 Funding:
- EPSRC iCASE studentship covers UK PhD tuition fees and provides a tax-free stipend (* £18,622 in 2023/24) for 4 years.
- Requires passing a Baseline Personnel Security Standard (BPSS) check.
- Full-time study only.
- 📚 Entry Requirements:
- Applicants need a minimum 2:1 degree in Computer Science.
- Prior ML/AI experience (education or work) desirable.
- Possession of a relevant taught MSc or master’s by Research advantageous.
- Available for full-time study.
- 🗓️ Thesis Submission:
- Expected within 48 months from the start of the PhD scholarship.
EPSRC iCASE Studentship – Assuring Large Language Models Through Knowledge Graph Infusion
Designation: PhD Researcher in Computer Science
Research Area: Neuro-Symbolic AI, Artificial Intelligence Assurance, Knowledge Graphs, Large Language Models
Location: Responsible Artificial Intelligence Group, University of Hull, UK
Eligibility/Qualification: Prospective candidates should:
- Hold or anticipate obtaining a relevant degree in Computer Science or a related field.
- Demonstrate a keen interest in the assurance and improvement of Large Language Models.
- Possess strong organizational and time management skills.
- Exhibit motivation to actively engage in the research program and contribute to the development of research questions.
- Have the ability to attend meetings, review literature, complete pertinent studies, and disseminate work at appropriate forums.
Job Description: As a PhD student, you will:
- Actively participate in the program of study, working on and developing research questions outlined in the proposal.
- Explore and implement Neuro-Symbolic AI approaches to detect hallucinations, falsehoods, contradictions, and knowledge gaps in LLM outputs.
- Contribute to the enhancement of LLM inference, training, and architecture for improved model quality.
- Engage with the Responsible Artificial Intelligence Group at the University of Hull, benefiting from the Postgraduate Training Scheme.
- Collaborate with QinetiQ and Naimuri for industrial supervision and support.
How to Apply: To apply, submit a personal statement addressing your motivation for pursuing PhD study, interest in the project, alignment of your skillset with project requirements, and the broader significance of research in this area. Visit Apply for the scholarship for detailed instructions.
Last Date for Apply: 20th March 2024
Interviews for shortlisted candidates will be conducted as part of the application process, and the successful candidate is expected to start in May 2024.
Join this cutting-edge research opportunity, contribute to the evolving field of AI assurance, and shape the future of Large Language Models. Apply now to be part of a dynamic research community at the forefront of Neuro-Symbolic AI.
