At VantaFold, the focus is on designing highly efficient deep learning methods and architectures, exploring ways to match the capabilities of much heavier counterparts. This aims to ensure that new student researchers can rely on innovative architectures and methods, rather than being gated by the need for extreme computing power.
Beyond efficiency, there is a strong interest in "fun ML" building creative and unorthodox applications of machine learning while keeping the computational footprint minimal.
Currently, VantaFold is an individual initiative. It represents an approach to start structuring ideas and experiments. There are no guarantees on specific outcomes, it's a challenge to see how these methods hold up in practice. The long-term goal is to build a foundation that might one day grow into collaborative research projects.