Computer Science Researcher & Web Developer
Specializing in NLP, LLM, and Computational Modeling
View My ProjectsI am a Computer Science researcher with a focus on NLP, LLM, and computational modeling. My research aims to bridge the gap in language technology for low-resource languages, particularly my native Burushaski language.
As a web developer, I'm seeking a position with Next Generation Web Development where I can apply my advanced knowledge of web development with my leadership abilities to meet our beloved, valued client's needs and their desired expectations.
Through which I will enhance the productivity of the company which associated with me.
As a Computer Science researcher, this is truly a dream come true. My first research paper (preprint) is now publicly available on arXiv.org.
I had the privilege of contributing to the Multilingual Physical Reasoning Benchmark Dataset by creating and submitting the Burushaski benchmark dataset, proudly representing one of the most unique and low-resource languages of Gilgit-Baltistan, Pakistan.
This research is part of the 5th Multilingual Representation Learning (MRL) Workshop, a global collaboration involving 335 researchers from 65 countries.
Abstract (summary): Global PIQA is a participatory commonsense reasoning benchmark covering over 100 languages and cultures. It includes 116 language varieties across five continents, 14 language families, and 23 writing systems. More than half of the examples reflect local foods, customs, and traditions. While state-of-the-art language models perform well overall, they show up to a 37% accuracy gap in low-resource languages. Global PIQA highlights that everyday cultural knowledge remains a key area for improvement in LLMs, beyond complex reasoning and expert knowledge.
I sincerely thank Adv. Wazir Shafi Lala, Quwáț K. Sunny Lala, and Jamsheed Aly Figaar for their invaluable support during the Burushaski dataset scripting and development process.
Special thanks to the lead authors Tyler A. Chang (University of California) and Catherine Arnett (EleutherAI) for their exceptional coordination, mentorship, and leadership throughout the Global PIQA initiative.
Feel free to reach out to me for research collaborations, web development projects, or discussions about NLP and low-resource language technology.
