Talk 1 

Yu

Topological edge states in two-dimensional materials and layered heterostructures

Recently, there has been active research development in two-dimensional materials beyond graphene. Two-dimensional networks containing transition metal elements exhibit fascinating physical properties with potential applications for field-effect transistors, spintronics, valleytronics, and thermoelectrics. Here we show that several two-dimensional materials consisting of a metal-organic framework, transition-metal chalcogenides, and halides have exciting features in their electronic band structure, which lead to novel magnetic interactions and topological characteristics like Chern insulators. From density-functional-theory calculations, we demonstrate that these compounds become ferromagnetic insulators with a non-trivial Chern number. While transition metal atoms are responsible for the ferromagnetic ground state, the band topology depends on the hopping matrix elements through chalcogen atoms. The nontrivial band topology is confirmed to have a nonzero Chern number, quantized Hall conductivity, and chiral edge states by using the Wannier function analysis. We propose strong dpσ-hybridization as a new channel of magnetic interactions in transition-metal chalcogenides materials. Further, we discuss possible manipulations of topological edge states in a (111)-oriented perovskite oxide heterostructures by incorporating spin-orbit coupling in a way that the neighboring perovskite oxide unit can be tuned by the size and character of the spin-orbit gaps.

Talk 2 

Song

Modeling of complex system with machine learning

In this talk, I would like to introduce extended work initiated from my APCTP period in JRG 'Designing principles of cellular network' leading by prof. Junghyo Jo. The group had been contributed to modeling of diverse complex systems. As a postdoc of the group, I was working on modeling of pancreatic islets which maintaining glucose homeostasis. I would like to briefly introduce biophysical model of pancreatic islets, and current my research topic regarding the model promoted by machine learning. 

Talk 3

Hiraoka

Burstiness in temporal networks

Non-Poissonian, so-called bursty temporal patterns are observed in a broad range of physical, biological, and social systems. In this talk, I will talk about two topics on how this burstiness is realized and what impact it has on the dynamical processes in complex networks. First, we look into the reconciliation of bursty dynamics of the activity of nodes and of interaction between nodes in communication networks. In order to explain this phenomenon, we introduce an activity-based temporal network model and show that it leads to heavy-tailed inter-event time distributions for both node and link activities. Second, we discuss how the burstiness affects the spreading phenomena in networks. We focus on the correlation between consecutive inter-event times and show that it leads to a slower spreading when the transmission involves more than one contact between nodes.

Talk 4 

Jo

Bursty time series analysis

Characterizing bursty temporal interaction patterns in complex systems is crucial to investigate the evolution and dynamics of such systems. The temporal interaction patterns have been described by a series of interaction events or event sequences, often showing non-Poissonian or bursty nature. Such bursty event sequences can be understood not only by heterogeneous inter-event times (IETs) but also by correlations between IETs. The heterogeneities of IETs have been extensively studied in recent years, while the correlations between IETs are far from being fully explored. In this talk, I introduce various measures for bursty time series analysis, such as the IET distribution, the burstiness parameter, the memory coefficient, the bursty train sizes, and the autocorrelation function, to discuss the relation between those measures. Then I show that the correlations between IETs can affect the dynamical behavior in complex systems, e.g., the speed of spreading taking place in temporal networks.

Talk 11 

Bayat

Measurement quench in many-body systems

Measurement is one of the key concepts which discriminates classical and quantum physics. Unlike classical systems, a measurement on a quantum system typically alters it drastically as a result of wave function collapse. Here we suggest that this feature can be exploited for inducing quench dynamics in a many-body system while leaving its Hamiltonian unchanged. Importantly, by doing away with dedicated macroscopic devices for inducing a quench—using instead the indispensable measurement apparatus only—the protocol is expected to be easier to implement and more resilient against decoherence. By way of various case studies, we show that our scheme also has decisive advantages beyond reducing decoherence—for spectroscopy purposes and probing nonequilibrium scaling of critical and quantum impurity many-body systems. 

Talk 12 

Mishra

Dynamics of many-body systems across the equilibrium quantum phase transitions

In this talk, I will address the non-equilibrium evolution of closed many-body quantum systems across the equilibrium quantum phase transition. We pay attention to the cases where the evolved state shows characteristic different behavior depending upon whether the equilibrium quantum phase transition is crossed or not. We focus on the evolution of entanglement in the XYZ spin chain and the revival of dynamical fidelity (Loschmidt echo) in long-range Kitaev wire and find that their survival is connected to the crossing of the equilibrium critical point. Finally, we consider the Ising model under periodic driving and explore that such system can be used for quantum metrology if the driving starts from the state close to the critical point. As the systems under consideration can be simulated in current experimental setups of ion traps and optical lattices, the present results are very much experimentally relevant.

Talk 13 

Park

Natural gradient descent for machine learning ground states

The natural gradient descent is a second-order optimization method introduced for machine learning (ML) purposes. This method utilizes local information geometry measured by the Fisher information matrix to deform the loss landscape. Thus it enables one to obtain better optimizing direction beyond the vanilla gradient descent. Even though several studies reported that the natural gradient descent works much better than first-order optimization methods, it is not commonly used for modern deep networks as it requires the inverse of a large matrix, i.e., inefficient in computation cost.

In this talk, we revisit the natural gradient descent for machine learning inspired variational quantum Monte-Carlo (VQMC). Unlike classical machine learning applications, the second-order method is the standard in VQMC as the usual first-order optimization methods do not work well. We introduce a possible reason why modern first-order optimization methods such as RMSProp and Adam do not work. Our argument is based on the viewpoint that modern first-order optimization methods employ the diagonal approximation of the Fisher information matrix. By comparing the property of the Fisher information matrix obtained from VQMC to that from real-world data (such as MNIST and CIFAR), we suggest that diagonal approximation only works in a particular case. We also discuss more conservative approximation methods proposed in ML context that can speed up VQMC without harming the accuracy.

Talk 14 

Mei

Matrix Product Solution of the Stationary State of Two-Species Open Zero Range Processes

Using the matrix product ansatz, we obtain solutions of the steady-state distribution of the two-species open one-dimensional zero range process. Our solution is based on a conventionally employed constraint on the hop rates, which eventually allows us to simplify the constituent matrices of the ansatz. It is shown that the matrix at each site is given by the tensor product of two sets of matrices and the steady-state distribution assumes an inhomogeneous factorized form. Our method can be generalized to the cases of more than two species of particles.

Talk 15 

Park

Data Science for the LHC physics

Talk 16 

Kuroyanagi

Gravitational waves: A new window to observe the Universe

Talk 17 

Terada

Primordial Black Holes: Implications for Dark Matter and Gravitational Waves

Talk 18 

Wong

An Independent Measurement of H0 from Lensed Quasars

Strong gravitational lens systems with time delays between the multiple images are a powerful probe of cosmology, particularly of the Hubble constant (H0) that is key to probing dark energy, neutrino physics, and the spatial curvature of the Universe, as well as discovering new physics.  The latest results from a total of six lenses constrains H0 to be 73.3(-1.8,+1.7) km/s/Mpc for a flat Lambda CDM cosmology, which is a measurement to 2.4% precision. These results are consistent with independent determinations of H0 using type Ia supernovae calibrated by the distance ladder method, and are in 3.1-sigma tension with the results of Planck CMB measurements.  Combined with the latest distance ladder results from the SH0ES project, we find a 5.3-sigma tension between Planck and late-Universe probes, hinting at possible new physics beyond the standard LCDM model and highlighting the importance of this independent probe.

Talk 19 

Hossain

Scalar field cosmology after Swampland

Swampland conjectures rule out the possibility of having de Sitter solution in string theory. This has severe impact on cosmology as this rules out the cosmological constant, the most favoured dark energy candidate. Now, observations tell us that from the recent past universe is expanding with acceleration which demands a de Sitter or quasi de Sitter solution. So observations and the swampland conjectures contradicting each other. In this talk, I will talk about the possibilities where we can evade this contradiction.

Talk 20 

Yeom

Annihilation to nothing: quantum gravitational wave function for inside a black hole

The interior of a static Schwarzschild metric can be written in terms of two functions similar to some models of anisotropic cosmology. With a suitable choice of canonical variables, we solve the Wheeler-DeWitt equation (WDW) inside the horizon of a Schwarzschild black hole. By imposing classicality near the horizon, and requiring boundedness of the wave function, we get a rather generic solution of the WDW equation, whose steepest-descent coincides well with the classical trajectory. However, there is an ambiguity in defining the arrow of time which leads to two possible interpretations – (i) If there is only one arrow of time, one can infer that the steepest-descent of the wave function follows the classical trajectory throughout, coming from the event horizon and going all the way down to the singularity, while (ii) if there were two arrows of time, it can be inferred that the steepest-descent of one of the wave functions comes inwards from the event horizon, and the other moves outwards from the singularity, and there exists an annihilation process of these two wave functions inside the horizon. Adopting the second interpretation could also shed some light on the information loss paradox: as time goes on, probabilities for histories that include black holes and singularities decay to zero and eventually only trivial geometries will dominate.