|
How well can large language models perform literature reviews?Intern: Abdullah Al KalbaniDegree course: BSc Computer Science LinkedIn Profile I am a final-year student. I chose Computer Science because I was fascinated by how quickly AI was developing and the impact it was beginning to have across different industries. Seeing the rapid progress in machine learning and AI technologies motivated me to learn more about how these systems work and how they can be used to solve real-world problems. During my degree, I have particularly enjoyed projects involving model training, AI evaluation, and applied machine learning systems. I am especially interested in building and evaluating AI systems to better understand their capabilities, limitations, and real-world applications. This interest is what drew me to this project on large language models and literature reviews. Why did you apply for this internship?I applied because it offered the opportunity to apply the evaluation-focused approach I developed during my third-year project, where I built a benchmark for comparing deepfake detection models, to assessing how large language models perform in literature reviews and understanding their strengths and limitations in realistic settings.The project also stood out to me because literature reviews are a major part of research, and LLMs are now being used for tasks like finding papers, summarising studies, and identifying research gaps. I wanted to explore where these tools can genuinely help, and where they still need careful human judgement. What do you hope to gain in completing this project?I hope to gain more experience in designing and carrying out a research project from start to finish, especially one that involves evaluating AI systems in a structured way.I want to improve my understanding of how LLMs can be used in academic research, beyond just producing summaries. By the end of the project, I hope to have a clearer idea of where these tools are genuinely useful, where they fail, and how their outputs can be evaluated properly. I also hope to develop my skills in literature review methods, research writing, and presenting findings clearly.
|
|
Sustainable AI Data Compression Case Study: Profiling BOA on BEDEIntern: Hanzila HussainDegree course: Bsc. Computer science with Industrial placement LinkedIn Profile What I find most compelling about Computer science is not just the technical depth, but the creative thinking it demands. Every problem is a puzzle waiting to be unpacked, and there is rarely just one way to solve it. That love of problem-solving is what drives me, whether I am working through an algorithmic challenge or designing a system architecture from scratch. Outside my studies, I spend my time running and bouldering. Both sports suit me well: running gives me space to think, while bouldering is, in many ways, just another form of problem-solving, figuring out a route one move at a time. Why did you apply for this internship?I applied because I want to move toward the research side of high-performance computing. Having worked with Transformers through my NLP modules, I am keen to explore the efficiency gains that State Space Models can offer. I also care about how AI is built, not just what it can do, and this project reflects the importance of prioritising approaches that are computationally considered and transparent, rather than simply throwing resources at a problem.What do you hope to gain in completing this project?I hope to gain hands-on experience working at the frontier of machine learning and high-performance computing, the kind of technical depth that coursework alone cannot replicate. Beyond the technical skills, I want to develop a sharper eye for research: learning how to frame problems rigorously, evaluate approaches critically, and push past the obvious solution.I am also excited to bring my existing knowledge into a real setting, particularly what I have built up around deep learning and systems, and see how it holds up against genuine research challenges. Most of all, I want to deliver something I am proud of: a project that makes a meaningful contribution to the work being done here, not just something that ticks a box.
|
|
The Language of Early Capitalism: A Computational AnalysisIntern: William RoperDegree course: BSc. Computer Science |
|
Identification of sub-groups of chronic obstructive pulmonary disease patients using clustering analysis of small airway histopathologyIntern: Urbi RoychowdhuryDegree course: BSc Mathematics and Statistics LinkedIn Profile I am a final-year student with a particular interest in statistical modelling, machine learning, and health data science. I originally chose mathematics because I loved problem-solving and understanding complex systems, but throughout my degree I became increasingly interested in statistics and working with real-world data. What I enjoy most is taking large, messy datasets and uncovering patterns that can help answer meaningful questions. Over the last few years, I have worked on projects ranging from machine learning applied to EEG data for schizophrenia classification to statistical modelling of Arctic sea-ice dynamics. Through these experiences, I discovered that I am especially drawn to health-related research because of its potential to make a tangible impact on people's lives. Outside my degree, I enjoy sewing, reading, running a university book club, and spending time on creative projects. I am particularly interested in the intersection of analytical thinking and creativity, and I enjoy finding opportunities to combine both! Why did you apply for this internship?I applied because it brings together several areas that I am particularly interested in: statistical modelling, machine learning, and health data science. One of the things that attracted me most to the project was the challenge of working with complex real-world clinical data. I am fascinated by how statistical methods can be used to uncover patterns that are not immediately obvious and help us better understand disease. I saw this internship as an opportunity to gain hands-on experience in health data research and to learn from researchers working in an area that I hope to pursue further in the future and take ownership of a complete health data research project and everything that comes with it! What do you hope to gain in completing this project?I hope to gain experience in taking ownership of a research project from start to finish. While I have previously worked with real-world datasets and completed research projects as part of my degree, this will be my first opportunity to lead a health data science project of this scale. I am particularly looking forward to developing my understanding of clustering methods, predictive modelling, and the challenges that come with analysing real clinical data. I am also excited to learn more about the research process itself: how questions are developed, how analytical decisions are made, and how findings are communicated. Beyond the technical skills, I am looking forward to becoming part of a research community. The opportunity to attend discussions, learn from researchers and postgraduate students, and present my work at the end of the internship is something I am especially excited about. Ultimately, I hope the project will help me grow in confidence as a researcher and deepen my understanding of how data science can contribute to improving healthcare.
|