Hello! I’m Thomas Helfer, and my academic and professional path has been driven by fascination with exploring new topics in science. My journey began at Imperial College London, where I delved into the intense world of theoretical physics during my master’s program. Transitioning from theoretical to computational physics for my Ph.D. was exciting. Having started with very little experience in coding, I was quickly able to upstart and get into the deep end of highly parallelized code in C++, even able to contribute to state-of-art codebases like GRChombo (now GRTL). My Ph.D. was a blend of theoretical concepts and practical simulations, offering me a comprehensive insight into the world of computational physics.

My first postdoc position marked another significant turn in my career. I became increasingly interested by deep learning, particularly superresolution. The remarkable achievements of DLSS in video games demonstrated to me the potential of these methods in scientific research, particularly in black hole simulations. This realization steered me towards deep learning, leading to my current role at the Institute for Advanced Computational Science at Stony Brook University. At Stony Brook, my focus is on leveraging deep learning to enhance simulations, exploring its applications not just in speeding up computational processes but also in broadening the scope of deep learning in scientific research. In addition to this, I have explored self-supervised learning (SSL) models and published a paper on multimodal learning using contrastive learning, further expanding the applications of AI in scientific research. What I love about my research is the opportunity to merge my background in physics with the exciting possibilities of AI, continually learning and applying these innovative techniques to new scientific questions.