A. F.
Research focus
My independent research focuses on AI safety, alignment, and the evaluation of multilingual language models. I investigate reliability gaps in how language models process identity-related information across different languages, scripts, and code-switched contexts. My work aims to develop empirical evaluation methods for identifying model failures and understanding their implications for safe and reliable AI deployment. I am also interested in model behavior, evaluation design, and culturally and linguistically robust AI systems.
Open to collaboration with independent researchers and volunteers interested in AI safety, multilingual LLM evaluation, model reliability, and evaluation methodology. I welcome research discussions, feedback on experimental design, dataset development, reproducibility checks, and joint research projects. Open to exploring other collaboration arrangements as suitable opportunities arise.
Papers on Damaqu (4)
- Multilingual Voice AI for Pilgrim Assistance: STT Reliability, Bias by Invisibility, and Ethical Deployment in Sacred Environmentsother
- Designing Culturally and Spiritually Sensitive AI Assistants for High-Density Religious Environments: A Prompt Engineering Framework Applied at Masjid al-Haramother
- AyeshaFareed/fatima-multilingual-identity-blind-spot · Datasets at Hugging Faceother
- MSG-Equivalent Analyzer: An Interactive Calculator for Predicting Naturally Occurring MSG Content in Culinary Spice Formulationsother