Shirin Dehghani

I am currently a Data Scientist at Metodata, specializing in computational social sciences. I focus on fine-tuning open and close source Large Language Models for diverse applications, optimizing advanced prompts, and developing AI-driven solutions, including LLM-based agents.

I hold a Master’s degree in Artificial Intelligence from Allameh Tabataba’i University, where my thesis explored the use of LLM embeddings for Persian stance detection and the development of a Persian benchmark, under the guidance of Dr. Mohammad Reza A. Oskoei.

My research interests span Natural Language Processing (NLP), Large Language Models (LLMs) and Artificial Intelligence. I am particularly interested in reliability and bias issues in AI, and the application of AI technologies in domains like social sciences and healthcare and etc.

I have also collaborated with the University of Padua on developing an innovative matrix completion method to enhance recommender systems, and I’m passionate about the ethical implications of datasets and AI technologies.

Email  /  CV  /  Google Scholar  /  X  /  Github  /  LinkedIn

My research interests are broad, Natural Language Processing (NLP), Large Language Models (LLMs), Deep Learning, Artificial Intelligence, Human-AI Interaction and Human-Computer Interaction.