Colloquia

Assistant Professor of Physics

University of Colorado – Boulder

Abstract: Generative models in machine learning have gained a lot of attention for their practical applications. In this talk, we explore how quantum correlations can enhance these models and investigate the potential quantum advantage based on the no-go theorem of hidden variable theory. Unlike other quantum machine learning theories that focus on sample complexity directly, our focus is on the expressive power of the learning models. We give an example showing how quantum correlations—specifically, contextuality—can define a quantum neural network that outperforms any reasonable classical neural network in terms of the number of hidden neurons required for a language translation task. This includes a proof for artificially constructed data and numerical results for real-world data. We will also briefly mention a possible mathematical framework that could solidify this claim. This direction is still in its early stages, and I hope this talk will inspire others to make more connections from foundational research in quantum information to practically useful problems in machine learning.

Bio: Xun Gao is an assistant professor at University of Colorado Boulder and an associate fellow at JILA. He got his PhD from Tsinghua University. Then he was a postdoc at Harvard University. His research interests are quantum computational advantage and quantum machine learning.

Assistant Professor of Physics

University of Colorado – Boulder

Abstract: Generative models in machine learning have gained a lot of attention for their practical applications. In this talk, we explore how quantum correlations can enhance these models and investigate the potential quantum advantage based on the no-go theorem of hidden variable theory. Unlike other quantum machine learning theories that focus on sample complexity directly, our focus is on the expressive power of the learning models. We give an example showing how quantum correlations—specifically, contextuality—can define a quantum neural network that outperforms any reasonable classical neural network in terms of the number of hidden neurons required for a language translation task. This includes a proof for artificially constructed data and numerical results for real-world data. We will also briefly mention a possible mathematical framework that could solidify this claim. This direction is still in its early stages, and I hope this talk will inspire others to make more connections from foundational research in quantum information to practically useful problems in machine learning.

Bio: Xun Gao is an assistant professor at University of Colorado Boulder and an associate fellow at JILA. He got his PhD from Tsinghua University. Then he was a postdoc at Harvard University. His research interests are quantum computational advantage and quantum machine learning.

Associate Research Professor

University of Colorado – Boulder, NREL

Abstract: Charge separation in organic photovoltaics appears to follow several mechanisms depending on the molecular details of the system in question. In this talk I present evidence from photoinduced absorption-detected magnetic resonance (PADMR) that charge transfer between phthalocyanine family donor molecules and a fullerene acceptor results in a driving force dependent distribution of charge-transfer distances, and that this behavior correlates well with the free charge generation yield measured via time-resolved microwave conductivity. We show that the highest driving force sample possesses both the lowest free charge yield and the shortest average charge transfer distance, as evinced by the relative strength of the magnetic dipole and isotropic exchange coupling. Conversely, the sample that displays optimal free charge yield at an intermediate driving force exhibits the weakest exchange coupling suggesting that the average charge transfer distance is substantially larger. These results are consistent with a charge separation model wherein a distribution of charge transfer distances governs whether the product state is free mobile carriers or a bound charge-transfer state. We suggest this as a parallel pathway that can allow free charge generation without the need for engineering specific intermolecular orientations, or necessarily including large amounts of disorder to drive charge separation.

Bio: Obadiah Reid earned a PhD in Physical Chemistry from the University of Washington in 2010. He joined NREL in 2010 and also held a research position at CU Boulder since 2014. His research interests include charge generation, transport, and recombination processes in organic semiconductors; development of new analytical tools and methods; and the convergent application of experimental and theoretical approaches.

Associate Research Professor

University of Colorado – Boulder, NREL

Abstract: Charge separation in organic photovoltaics appears to follow several mechanisms depending on the molecular details of the system in question. In this talk I present evidence from photoinduced absorption-detected magnetic resonance (PADMR) that charge transfer between phthalocyanine family donor molecules and a fullerene acceptor results in a driving force dependent distribution of charge-transfer distances, and that this behavior correlates well with the free charge generation yield measured via time-resolved microwave conductivity. We show that the highest driving force sample possesses both the lowest free charge yield and the shortest average charge transfer distance, as evinced by the relative strength of the magnetic dipole and isotropic exchange coupling. Conversely, the sample that displays optimal free charge yield at an intermediate driving force exhibits the weakest exchange coupling suggesting that the average charge transfer distance is substantially larger. These results are consistent with a charge separation model wherein a distribution of charge transfer distances governs whether the product state is free mobile carriers or a bound charge-transfer state. We suggest this as a parallel pathway that can allow free charge generation without the need for engineering specific intermolecular orientations, or necessarily including large amounts of disorder to drive charge separation.

Bio: Obadiah Reid earned a PhD in Physical Chemistry from the University of Washington in 2010. He joined NREL in 2010 and also held a research position at CU Boulder since 2014. His research interests include charge generation, transport, and recombination processes in organic semiconductors; development of new analytical tools and methods; and the convergent application of experimental and theoretical approaches.

Physics Professor

University of Colorado – Boulder

Abstract: Understanding and ultimately controlling the properties of matter, from molecular to quantum systems, requires imaging the elementary excitations on their natural time and length scales. To achieve this goal, we developed scanning probe microscopy with ultrafast and shaped laser pulse excitation for multiscale spatio-temporal optical nano-imaging. In corresponding ultrafast movies, we resolve the fundamental quantum dynamics from the few-femtosecond coherent to the thermal transport regime. I will discuss specific examples visualizing in space and time the nanoscale heterogeneity in competing structural and electronic dynamic processes in matter. In photovoltaic perovskites with far-from equilibrium excitation we visualize the polaron dynamics and its spatial, temporal and fluence dependence on the nanoscale, reflecting the elementary processes underling their photoresponse. In the extension to coherent nonlinear nanoimaging, we resolve competing ultrafast intra- and interlayer dynamic processes in graphene and 2D semiconductors and their heterostructures. As a perspective I will show that we are reaching the ultimate goal of functional imaging and control, to link macroscopic properties to microscopic interactions in materials at their fundamental spatio-temporal scales.

Bio: Markus Raschke is professor at the Department of Physics and JILA at the University of Colorado at Boulder. His research is on the development and application of nano-scale nonlinear and ultrafast spectroscopy to control the light-matter interaction on the nanoscale. These techniques allow for imaging structure and dynamics of molecular and quantum matter with nanometer spatial resolution. He received his PhD in 2000 from the Max-Planck Institute of Quantum Optics and the Technical University in Munich, Germany. Following research appointments at the University of California at Berkeley, and the Max-Born-Institute in Berlin, he became faculty member at the University of Washington in 2006, before moving to Boulder in 2010. He is fellow of the Optical Society of America, the American Physical Society, he American Association for the Advancement of Science, and the Explorers Club.

Physics Professor

University of Colorado – Boulder

Abstract: Understanding and ultimately controlling the properties of matter, from molecular to quantum systems, requires imaging the elementary excitations on their natural time and length scales. To achieve this goal, we developed scanning probe microscopy with ultrafast and shaped laser pulse excitation for multiscale spatio-temporal optical nano-imaging. In corresponding ultrafast movies, we resolve the fundamental quantum dynamics from the few-femtosecond coherent to the thermal transport regime. I will discuss specific examples visualizing in space and time the nanoscale heterogeneity in competing structural and electronic dynamic processes in matter. In photovoltaic perovskites with far-from equilibrium excitation we visualize the polaron dynamics and its spatial, temporal and fluence dependence on the nanoscale, reflecting the elementary processes underling their photoresponse. In the extension to coherent nonlinear nanoimaging, we resolve competing ultrafast intra- and interlayer dynamic processes in graphene and 2D semiconductors and their heterostructures. As a perspective I will show that we are reaching the ultimate goal of functional imaging and control, to link macroscopic properties to microscopic interactions in materials at their fundamental spatio-temporal scales.

Bio: Markus Raschke is professor at the Department of Physics and JILA at the University of Colorado at Boulder. His research is on the development and application of nano-scale nonlinear and ultrafast spectroscopy to control the light-matter interaction on the nanoscale. These techniques allow for imaging structure and dynamics of molecular and quantum matter with nanometer spatial resolution. He received his PhD in 2000 from the Max-Planck Institute of Quantum Optics and the Technical University in Munich, Germany. Following research appointments at the University of California at Berkeley, and the Max-Born-Institute in Berlin, he became faculty member at the University of Washington in 2006, before moving to Boulder in 2010. He is fellow of the Optical Society of America, the American Physical Society, he American Association for the Advancement of Science, and the Explorers Club.