Quantum Information Flash Talks
By Leon Guerrero
As a new year rolled around, I am sure many of us were keen to see the pandemic come to an end and be able to attend events in person. Unfortunately, as we entered Hillary term government restrictions forced many of us to remain at home and away from the city of dreaming spires. We were unable to host events in person, but if there is one thing we have gained from this pandemic, it is the ability to host informative and hopefully interesting online talks about quantum information. We kicked of this term with flash talks from three DPhil students currently performing their research at the Oxford University. These talks were given live on Zoom, streamed on our Facebook page, and in case you missed them have now been posted to our YouTube channel.
Cameron Booker, a mathematician at heart, appears to have gotten lost when he joined Oxford for his DPhil. He is currently researching the dynamics of out of equilibrium quantum many body systems. This may sound complicated, and that's because it is. That being said, Cameron gave a fantastic and accessible introduction to this very theoretical field using only two equations! He showed us how it is incredibly hard to solve the Schrodinger equation for many particle systems, even with the University's supercomputer, due to the curse of dimensionality. This curse only becomes worse (curse of dimensionality)2 when you move to the Lindblad equation that describes open quantum systems. This curse can be somewhat cured with the help of symmetries and allows you to solve certain problems relating to quantum synchronization and time crystals. This research is of huge interest for taming quantum mechanical systems that will be useful for developing quantum technologies.
Our next talk was given by Lewis Anderson, titled 'The dizzying heights and barren plateaus of variational quantum algorithms'. This is a relatively new topic that you probably won't find in your standard quantum computing textbook, thankfully Lewis gave a fantastic summary of the topic and recent developments in literature. Variational quantum algorithms are essentially quantum algorithms that solve optimization problems and have many potential applications in machine learning, quantum chemistry and non-linear partial differential equations. However, these algorithms encounter the problems of barren plateaus, where the algorithm takes too long to reach to desired answer and the gradients decrease exponentially as you increase the number of qubits. This can happen if the initial ansatz is too broad, and so similar to our last talk, the idea of symmetry can help! Problem specific arguments and symmetries help avoid barren plateaus by increasing the trainability of these algorithms to allow them to reach their peak potential.
Last but definitely not least was a talk from former QuInSoc president Maria Violaris. Taking a different approach to our other speakers, Maria, instead of using PowerPoint to present her ideas, just needed a bowl, a couple of eggs and a whisk. But she was not making a cake, instead she gave a fascinating talk about a quantum model for the arrow of time and how this model for irreversibility could be connected to the ultimate limits of erasing quantum information. Our classical model for the flow of time is given by the second law of thermodynamics, where the entropy of an isolated system is exceedingly unlikely to decrease as time progresses. The live example given by Maria was how you cannot unscramble an egg. This law is however just a statistical law that only works for large systems. Instead Maria suggested a model where the reversibility was defined as to whether you could transform the egg from one state to another. You still encounter the same indication of the arrow of time as you are unable to transform the mixed/scrambled state to a pure/separated state. She then related this new model to a quantum version of the famous Szilard's engine thought experiment, with the audience playing the role of Maxell's demon, and demonstrated how there is an additional fundamental energy cost for erasing information. Physicist Edwin James once said that information, quantum mechanics and entropy were scrambled up into an omelet that needs unscrambling and perhaps this new way of looking at irreversibility can help us unscramble this omelet.
My summaries given here do not do justice to the talks themselves and so I encourage anyone reading this to check out their talks at the link here.