Back

“It from bit” in the era of machine learning

The famous “It from Bit” slogan expresses a foundational research approach that emphasizes that all of our perceptions and concepts of the physical world are ultimately given in the form of data.

With the ongoing machine learning revolution, it is fruitful to investigate how the research of exotic foundational physics can profit from our new methods in data science. For this purpose, I will present two examples of a very different nature. The first example uses a Variational Auto-Encoder modification to discover disentangled factors of variation in data. The second example constructs a variational quantum circuit using effects of quantum field theory on curved spacetime. Using a variational quantum sensing strategy, it can be applied to distill information about the quantum field or the spacetime when no handcrafted protocol is known. 

Information

 

Speaker: Marius Krumm (University of Innsbruck)

Time:  14:00