Hillary N. Owusu

Hillary N. Owusu

PhD Student · CLIP Lab · University of Maryland
Email LinkedIn Google Scholar CV
looking for research opportunities · internships · collaborations

Hi! I am a third-year PhD student in Computer Science at the University of Maryland, advised by Prof. Naomi Feldman in the CLIP Lab.

My research focuses on AI Safety and Alignment: understanding how language models behave when they reason and when they are influenced. I use mechanistic interpretability and causal intervention to study model internals, drawing on cognitive science and computational social science to understand what those internals mean.

I also care about AI education, helping students and engineers engage with these systems critically. I coordinate ML programming at UMD and have mentored undergraduate researchers through full research cycles.

I am currently looking for research opportunities in language-model reliability, interpretability, and responsible AI, as well as collaborations on aligned research problems.

Publications
ACL 2026
Anchoring Depends on Confidence and Post-Training in Language Models
Hillary N. Owusu, Naomi H. Feldman
Shows that anchoring resistance tracks a model’s internal certainty more strongly than factual accuracy, and that post-training changes susceptibility across Llama and Qwen variants.
[paper]
arXiv 2026preprint
Localizing Anchoring Pathways in Language Models
Hillary N. Owusu, Sarah Wiegreffe, Naomi H. Feldman
Uses attribution-based circuit localization to identify anchor-sensitive pathways in 7B–8B Llama and Qwen models, finding stronger faithfulness for edge-level methods and changes in pathway importance after instruction tuning.
[paper]
arXiv 2025preprint
Bias-Aware AI Chatbot for Engineering Advising at the University of Maryland
et al., Hillary N. Owusu
RAG-based advising chatbot with integrated bias detection. Mentored four undergraduate researchers from project scoping through publication as senior author.
[paper]
Projects
Agentic AI · Research Automation · Hackathon
CouncilAgent / AnchorBreak
Team-built multi-agent research system that detects when an AI council has converged or stalled and forces a counter-design from a new hypothesis family. On a machine-unlearning benchmark, the system improved over baseline by 62.7% and reached 89% of the human-best score.
[project demo]
Robustness · Benchmarking
GSM++ Reasoning Robustness Framework
Stress-tested LLaMA reasoning across 5 semantic variants of GSM8K. Accuracy dropped from 71% to 55.5% under Hindi translation.
Privacy · Security
Membership Inference Attack on GPT-2
Built a TF-IDF attacker on a GPT-2 shadow model achieving 97% accuracy at distinguishing training members from non-members.
Fairness · NLP · Low-resource Languages
Gender Bias in English to Twi Neural Machine Translation
Audited gender bias in English-Twi NMT across 3,080 minimal pairs. Found systematic gendered semantic drift in translations.
Experience & Service
Jun 2026
Selected Participant, Machine Learning Summer School (MLSS) NYC 2026
Selected as one of approximately 200 participants from more than 1,200 applicants for the two-week program at Columbia University.
2026
Reviewer, NeurIPS 2026
2026
Reviewer, ACL 2026
2026
Reviewer, ICML Workshop on Technical AI Governance Research
2026
Reviewer, ICML AI4GOOD Workshop
2026
Mentor, HASTC Workshop, Responsible AI for Health Symposium, JHU
2025
Reviewer, NeurIPS 2025 Reliable ML Workshop
Jul–Aug 2025
Research Mentor, UMD ESTEEM/SER-Quest Summer Program
Mentored 4 undergraduates through the Bias-Aware Chatbot project; senior author on resulting preprint.
2025–now
ML Instructor & Coordinator, CMSE TLP Program, UMD
Teaching ML skills and AI applications to undergraduate engineers across disciplines.
2023–2024
Facilitator, CMSE Summer Bridge Program, UMD
2023–2024
Teaching Assistant, UMD
Algorithms (CMSC351), Data Science (CMSC320), C Programming (CMSC106)
Last updated July 2026 · Built with curiosity and care.