Mark Neumann
I'm a researcher and engineer interested in neural networks and interactive machine learning. Currently I'm a Principal Scientist at Evolutionary Scale (now part of Biohub), where I develop representation learning models for protein sequences and structures. Previously I was the Head of Machine Learning at Orbital Materials, where I led the development of foundation models for materials science.
Before switching to applying neural networks in the hard sciences, I worked on natural language processing at the Allen Institute for Artificial Intelligence. I cocreated AllenNLP, a library for Natural Language Processing research, focused on neural networks. I also worked on Elmo (Embeddings from Language Models) a popular representation learning method for embedding words in context. It began a trend of deep contextualised word representation learning in NLP, such as BERT/GPT-3.