Sarah Wyer

Sarah Wyer is a PhD student and Teaching Fellow in the Department of Computer Science at Durham University. Her research focuses on mitigating gender and social class bias in large language models through an intersectional lens. She is affiliated with the EquiAI UK Network and the Artificial Intelligence and Human Systems Group.

Education: PhD in Computer Science (Durham University). Thesis: 'Identifying and Mitigating Gender and Social Class Bias in Large Language Models.'

Research Interests: Bias in AI, meta-learning language models, AI ethics, widening participation in HE, and promoting diversity in STEM. She explores how algorithmic systems perpetuate social inequalities and advocates for equitable AI design.

Teaching Roles: Module convenor for Foundation Computer Science and teaches English for Scientists with Project. Her teaching emphasizes accessibility and inclusivity in STEM education.

Publications: Recent work includes studies on sample diversity in LLMs (2024) and gendered violence in GPT-3 outputs (2025). These publications highlight her focus on ethical AI and societal impacts.

Advising & Grants: Supervised by Prof Sue Black and Prof Nicola Whitton. Engaged in pedagogical innovations to support underrepresented groups in computing.

Labs/Teams: Active in the EquiAI UK Network and AI and Human Systems Group, collaborating on interdisciplinary projects to address systemic biases in technology.