Abstract for Talk 1:
We computationally studied the photoisomerization reaction of the retinal chromophore in rhodopsin using a two-state two-mode model coupled to thermal baths. Reaction quantum yields at the steady state (10 ps and beyond) were found to be considerably different than their transient values, suggesting a weak correlation between transient and steady-state dynamics in these systems. Significantly, the steady-state quantum yield was highly sensitive to minute changes in system parameters, while transient dynamics was nearly unaffected. Correlation of such sensitivity with standard level spacing statistics of the nonadiabatic vibronic system suggests a possible origin in quantum chaos. The significance of this observation of quantum yield parametric sensitivity in biological models of vision has profound conceptual and fundamental implications.
Abstract for Talk 2:
The use of near term quantum devices for compression of information is an exciting prospect which can enable the use of quantum resources for complex tasks. To this end, different compression algorithms, including the quantum autoencoder, have been proposed. These algorithms rely on trained parameterized quantum circuits to perform the compression. The success of the training depends on the structure of the employed circuit, whose design can be difficult to generalize. In this work we propose a novel strategy to design quantum circuits using an evolutionary algorithm, with a restricted gate set based on classical logic operations. The use of the limited gate set enables efficient simulation of the quantum circuit. We show initial applications for compression of different family of states, including single particle states, two particle states, random states, prime states, among others. This opens a new path for using near term quantum devices for compressing quantum data and facilitating efficient quantum simulations for various tasks.