Ron Eliav Homepape

I’m a Ph.D. Candidate in Computer Science at Bar-Ilan University, proudly advised by Prof. Ido Dagan.

My research focuses on advancing language models toward achieving a nuanced, soft understanding of language. This involves developing methods to significantly mitigate model hallucination and enhance the overall reliability of generative outputs. Furthermore, my work extends to improving the rigorous evaluation of model faithfulness, specifically by closing the gap between outputs that are merely plausible and those that are truly grounded in the source material.

As a complementary line of applied research, I’m passionate about applying NLP principles to education, aiming to create personalized and adaptive learning experiences. My ongoing project leverages NLP to tailor educational content (e.g., Math, English as a foreign language) to each student’s skills and understanding. I’m excited that this project is already making an impact -- it’s being used daily by thousands of high school students across Israel.

Before beginning my Ph.D. journey, I completed both my B.Sc. and M.Sc. in Computer Science at Bar-Ilan University, also under Prof. Dagan’s supervision. Along the way, I had the opportunity to conduct research during internships at Cornell University with Prof. Yoav Artzi and at IBM Research -- experiences that deeply shaped my perspective on the value of collaboration and learning from diverse research communities.

If you’re interested in discussing model hallucination, AI in education, or any other exciting area at the intersection of language models and human usefulness, feel free to reach out!