I'm an MPhil student at The Chinese University of Hong Kong working under Prof. Farzan Farnia. My research focuses on generative models, large language models, and optimization.
- Machine Learning Theory
- Optimization
- Large Language Models
- Statistics
- Generative Models
PromptSplit: Revealing Prompt-Level Disagreement in Generative Models
Mehdi Lotfian, Mohammad Jalali, Farzan Farnia
arXiv:2602.04009 [cs.LG]
arXiv • Code
- PromptSplit — Kernel-based framework for detecting prompt-dependent disagreements in text-to-image and LLM models (4 stars)
- Artificial-Intelligence-Mini-Projects — Classical algorithms, ML, and deep learning course projects
- Machine-Learning-Small-Projects — Implementations from graduate ML coursework
- Additional repositories (NLP, IR, sentiment analysis, etc.) available in my profile
Passionate about theoretically grounded generative AI.
Last updated: March 2026