Karun Sharma

I'm Karun, a research scholar working with advisors at Halmstad University and Meta on multimodal LLMs and vision-language models. I focus on scientific question generation and improving how models understand visual and textual information together, using PyTorch and CUDA for distributed training.

Before this I was a Computer Vision Engineer at Zocket and studied CS at Bennett University.

Ask me about vision-language models, scientific question generation in LLMs, PyTorch optimization, and multimodal AI research.