Winning entry for the 2017 CCIMI video contest. Video:
Many thanks to falcxne for allowing use of his songs “Said You Wouldn’t Change” (from the album Timbits III), “I Hate Homework (Interlude)” (from Timbits II) and “NYC (Interlude)” (from Timbits III); you can find these and more at https://falcxne.bandcamp.com/. And thanks to Phillipe Faist for creating the entropy zoo and allowing me to use it in the video. For more details on the continuity bounds mentioned, you can find a preprint here, which is joint work with my supervisor, Dr. Nilanjana Datta.
Lastly, one clarification: in the video, I talk about the data-compression example in the language of classical information theory, for simplicity. However, the quantum case can be described in a very similar way, with the von Neumann entropy replacing the Shannon entropy.
The slides for the talk I gave at the Autrans summer school Stochastic Methods in Quantum Mechanics on my work with Alain Joye, Yan Pautrat, and Renaud Raquepas on Landauer’s Principle in repeated interaction systems.
“Landauer’s principle states that the energy cost to erase one bit of information by the action of a thermal reservoir at equilibrium temperature $T$ is always at least $kT$, and provides an interesting link between information theory and physics. My goal is to discuss Landauer’s principle generally, and for the case of a 2 level system coupled to a thermal reservoir modelled by an XY spin chain (or sequence of spin-1⁄2 particles), using the tools of quantum statistical mechanics.”
“The promise problem QSEP-STATE asks if a quantum state described by a circuit is close to a separable (not entangled) state across a given cut, or not. A quantum multiprover protocol was found to give an upper bound for the computational complexity of QMIPne, which is the class of problems that can be solved by a computationally bounded quantum verifier exchanging messages with unentangled, computationally bounded, quantum provers.”
“QSEP-CIRCUIT, as defined in a recent paper by Patrick Hayden, Kevin Milner, and Mark Wilde (http://arxiv.org/abs/1211.6120) in the computer science department here at McGill, is the problem of determining whether or not a quantum state given as a circuit is separable or not. The problem has been shown to be in QIP(2), and to be NP-hard and QSZK-hard, which lends credence to the idea that the problem is QMA-hard. The goal of this talk is to provide background for and explain in detail what QSEP-CIRCUIT and QMA are, explain why we care whether or not QSEP-CIRCUIT is QMA-hard, and hopefully showcase quantum information theory as an exciting and dynamic field. No background knowledge will be assumed.”