doi.bio/sebastian_w_bodenstein
Sebastian W Bodenstein
Early Life and Education
Sebastian Bodenstein is a research engineer at DeepMind, focusing on protein folding. He previously worked in the Wolfram Research machine learning group, developing neural network frameworks and applications.
Bodenstein holds a PhD in theoretical physics from the University of Cape Town, which he completed in 2014. He also obtained a BSc in Mathematics and Physics from the same university in 2007.
Career
Bodenstein joined DeepMind as a research engineer in 2020. Prior to that, he worked at Wolfram Research, where he was involved in neural network framework development and contributed to MXNet. He is one of the two main creators of the Mathematica Neural Net Framework and led the development of the Wolfram Neural Net Repository.
Bodenstein has a background in theoretical physics and made the transition into machine learning in 2013, following the advent of AlexNet. He has also worked on reinforcement learning tutorials, such as the Indaba Reinforcement Learning Tutorial, and various other projects, including MongoLink and the PC Causal Discovery Algorithm.
Publications
Bodenstein has published several articles, including:
- Solutions to Penrose's The Road to Reality
- Highly accurate protein structure prediction with AlphaFold
- AlphaZero: A landmark result in Artificial Intelligence research, where a single algorithm mastered Chess, Go, and Shogi
Sebastian W Bodenstein
Early Life and Education
Sebastian Bodenstein is a research engineer at DeepMind, focusing on protein folding. He previously worked in the Wolfram Research machine learning group, where he led the development of the Wolfram Neural Net Repository.
Bodenstein holds a PhD in theoretical physics from the University of Cape Town (2014) and a BSc in Mathematics and Physics (2007) from the same institution.
Career
Before entering the field of machine learning, Bodenstein did a PhD in theoretical physics. He is one of the two main creators of the Mathematica Neural Net Framework and has contributed to MXNet. He also designed a reinforcement learning tutorial, Indaba, for a wide audience.
Bodenstein joined DeepMind as a research engineer in 2020. His work at DeepMind focuses on protein folding, unsupervised learning, and reinforcement learning.
Publications
- MongoLink: a package for interacting with MongoDB inside the Wolfram Language via the high-performance MongoDB C driver.
- The PC Causal Discovery Algorithm: a Mathematica implementation.
- Solutions to Penrose’s The Road to Reality.
Sebastian W Bodenstein
Early Life and Education
Sebastian Bodenstein is a research engineer at DeepMind, focusing on protein folding. He previously worked in the Wolfram Research machine learning group, developing neural network frameworks and applications.
Bodenstein holds a PhD in theoretical physics from the University of Cape Town, which he completed in 2014. He also obtained a BSc in Mathematics and Physics from the same university in 2007.
Career
Before entering the field of machine learning, Bodenstein did a PhD in theoretical physics. He joined DeepMind as a research engineer in 2020. He is also an MXNet contributor.
Bodenstein has been involved in several notable projects, including:
- The Mathematica Neural Net Framework, where he was one of the two main creators.
- The Wolfram Neural Net Repository, where he led the development team.
- Designing a reinforcement learning tutorial (Indaba Reinforcement Learning Tutorial) for a wide audience.
- Working on the MongoLink package for interacting with MongoDB inside the Wolfram Language.
- Providing solutions to Penrose's The Road to Reality.
Publications
Bodenstein has published work on highly accurate protein structure prediction with AlphaFold. He has also written about AlphaZero, a single algorithm that has mastered Chess, Go, and Shogi. AlphaZero is the successor to AlphaGo, which beat a champion Go player without requiring human games, unlike its predecessor.