Top arXiv papers

We present the discovery of a Neptune-mass planet orbiting a 0.8 +- 0.3 M_Sun star in the Galactic bulge. The planet manifested itself during the microlensing event MOA 2011-BLG-028/OGLE-2011-BLG-0203 as a low-mass companion to the lens star. The analysis of the light curve provides the measurement of the mass ratio: (1.2 +- 0.2) x 10^-4, which indicates the mass of the planet to be 12-60 Earth masses. The lensing system is located at 7.3 +- 0.7 kpc away from the Earth near the direction to Baade's Window. The projected separation of the planet, at the time of the microlensing event, was 3.1-5.2 AU. Although the "microlens parallax" effect is not detected in the light curve of this event, preventing the actual mass measurement, the uncertainties of mass and distance estimation are narrowed by the measurement of the source star proper motion on the OGLE-III images spanning eight years, and by the low amount of blended light seen, proving that the host star cannot be too bright and massive. We also discuss the inclusion of undetected parallax and orbital motion effects into the models, and their influence onto the final physical parameters estimates.

We classify inequivalent binary extended $1$-perfect bitrades of length $10$, constant-weight extended $1$-perfect bitrades of length $12$, and STS bitrades derived from them.

The representation category $\mathcal{A} = Rep(G,\epsilon)$ of a supergroup scheme $G$ has a largest proper tensor ideal, the ideal $\mathcal{N}$ of negligible morphisms. If we divide $\mathcal{A}$ by $\mathcal{N}$ we get the semisimple representation category of a pro-reductive supergroup scheme $G^{red}$. We list some of its properties and determine $G^{red}$ in the case $Gl(m|1)$.

We present three major transitions that occur on the way to the elaborate and diverse societies of the modern era. Our account links the worlds of social animals such as pigtail macaques and monk parakeets to examples from human history, including 18th Century London and the contemporary online phenomenon of Wikipedia. From the first awareness and use of group-level social facts to the emergence of norms and their self-assembly into normative bundles, each transition represents a new relationship between the individual and the group. At the center of this relationship is the use of coarse-grained information gained via lossy compression. The role of top-down causation in the origin of society parallels that conjectured to occur in the origin and evolution of life itself.

Here we show how to determine the orbital parameters of a system composed of a star and N companions (that can be planets, brown-dwarfs or other stars), using a simple Fourier analysis of the radial velocity data of the star. This method supposes that all objects in the system follow keplerian orbits around the star and gives better results for a large number of observational points. The orbital parameters may present some errors, but they are an excellent starting point for the traditional minimization methods such as the Levenberg-Marquardt algorithms.

We present a scheme for generating robust steady-state entanglement between qubits that do not interact and do not exchange information, using Markovian reservoir engineering. Several realizations including natural or artificial atoms coupled to a single cavity or to separate cavities with a uni- or bidirectional decay channel between them are considered. The scheme requires no dynamic control, measurements or feedback, and is optimized to be robust with regard to arbitrarily large cavity and fiber losses, thereby opening a route to stabilization of truly long-range entanglement between distant qubits. The protocol is also robust with regard to uncertainty in the system parameters and atomic decoherence and requires no knowledge of the initial state of the system.

The compact X-ray source in the eclipsing X-ray binary IC 10 X-1 has reigned for years as ostensibly the most massive stellar-mass black hole, with a mass estimated to be about twice that of its closest rival. However, striking results presented recently by Laycock et al. reveal that the mass estimate, based on emission-line velocities, is unreliable and that the mass of the X-ray source is essentially unconstrained. Using Chandra and NuSTAR data, we rule against a neutron-star model and conclude that IC 10 X-1 contains a black hole. The eclipse duration of IC 10 X-1 is shorter and its depth shallower at higher energies, an effect consistent with the X-ray emission being obscured during eclipse by a Compton-thick core of a dense wind. The spectrum is strongly disk-dominated, which allows us to constrain the spin of the black hole via X-ray continuum fitting. Three other wind-fed black-hole systems are known; the masses and spins of their black holes are high: M ~ 10-15 Msun and a*>0.8. If the mass of IC 10 X-1's black hole is comparable, then its spin is likewise high.

We propose a theory of interference contributions to the two-dimensional exciton diffusion coefficient. The theory takes into account four spin states of the heavy-hole exciton. An interplay of the single particle, electron and hole, spin splittings with the electron-hole exchange interaction gives rise to either localization or antilocalization behavior of excitons depending on the system parameters. Possible experimental manifestations of exciton interference are discussed.

Compact radio sources sometimes exhibit intervals of large, rapid changes in their flux-density, due to lensing by interstellar plasma crossing the line-of-sight. A novel survey program has made it possible to discover these "Extreme Scattering Events" (ESEs) in real time, resulting in a high-quality dynamic spectrum of an ESE observed in PKS 1939-315. Here we present a method for determining the column-density profile of a plasma lens, given only the dynamic radio spectrum of the lensed source, under the assumption that the lens is either axisymmetric or totally anisotropic. Our technique relies on the known, strong frequency dependence of the plasma refractive index in order to determine how points in the dynamic spectrum map to positions on the lens. We apply our method to high-frequency (4.2-10.8 GHz) data from the Australia Telescope Compact Array of the PKS 1939-315 ESE. The derived electron column-density profiles are very similar for the two geometries we consider, and both yield a good visual match to the data. However, the fit residuals are substantially above the noise level, and deficiencies are evident when we compare the predictions of our model to lower-frequency (1.6-3.1 GHz) data on the same ESE, thus motivating future development of more sophisticated inversion techniques.

Every closed oriented manifold $M$ is associated with a set of integers $D(M)$, the set of self-mapping degrees of $M$. In this paper we investigate whether a product $M\times N$ admits a self-map of degree $d$, when neither $D(M)$ nor $D(N)$ contains $d$. We find sufficient conditions so that $d\notin D(M\times N)$, obtaining, in particular, products that do not admit self-maps of degree $-1$ (strongly chiral), that have finite sets of self-mapping degrees (inflexible) and that do not admit any self-map of degree $dp$ for a prime number $p$. Furthermore we obtain a characterization of odd-dimensional strongly chiral hyperbolic manifolds in terms of self-mapping degrees of their products.

Let $\mathcal{T}(\mathcal{N})$ be a nest algebra of operators on Hilbert space and let $\mathcal{L}$ be a weakly closed Lie $\mathcal{T}(\mathcal{N})$-module. We construct explicitly the largest possible weakly closed $\mathcal{T}(\mathcal{N})$-bimodule $\mathcal{J}(\mathcal{L})$ and a weakly closed $\mathcal{T}(\mathcal{N})$-bimodule $\mathcal{K}(\mathcal{L})$ such that \[ \mathcalJ(\mathcalL)⊆\mathcalL ⊆\mathcalK(\mathcalL) +\mathcalD_\mathcalK(\mathcalL), \] $[\mathcal{K}(\mathcal{L}), \mathcal{T}(\mathcal{N})]\subseteq \mathcal{L}$ and $\mathcal{D}_{\mathcal{K}(\mathcal{L})}$ is a von Neumann subalgebra of the diagonal $\mathcal{T}(\mathcal{N})\cap \mathcal{T}(\mathcal{N})^*$.

Our renormalization group consistent variant of optimized perturbation, RGOPT, is used to calculate the nonperturbative QCD spectral density of the Dirac operator and the related chiral quark condensate $\langle \bar q q \rangle$, for $n_f=2$ and $n_f=3$ massless quarks. Sequences of approximations at two-, three-, and four-loop orders are very stable and give $\langle \bar q q \rangle^{1/3}_{n_f=2}(2\, {\rm GeV}) = -(0.833-0.845) \bar\Lambda_2 $, and $ \langle \bar q q \rangle^{1/3}_{n_f=3}(2\, {\rm GeV}) = -(0.814-0.838) \bar\Lambda_3 $ where the range is our estimated theoretical error and $\bar\Lambda_{n_f}$ the basic QCD scale in the $\rm \bar{MS}$-scheme. We compare those results with other recent determinations (from lattice calculations and spectral sum rules).

Current-induced domain wall (DW) motion has drawn great attention in the last decades as the key operational principle of emerging magnetic memory devices. As the major driving force of the current-induced DW motion, the spin-orbit torque (SOT) on chiral DWs has been proposed and extensively studied nowadays. However, we demonstrate here that there exists another driving force, which is larger than the SOT in ultra-thin Co films. Moreover, the direction of the present force is found to be opposite to the prediction of the spin-transfer torque (STT), resulting in the DW motion along the current direction. The symmetry of the force and its peculiar dependence on the DW structure suggest that the present force is, most likely, attributed to considerable enhancement of the nonadiabatic STT with a negative spin polarization in atomically thin Co layers. These findings open a new pathway to enhance the overall spin torque efficiency-the critical parameter in emerging spintronic devices.

We present a detailed analysis of the cosmic string spectrum. Explicit solutions are numerically found using Mathematica and presented here for the lowest-lying supported modes. Most of the emphasis is on the Nambu-Goldstone modes and the least massive excitation, the latter of which is shown to be the scalar breather mode. We address the possibility of pseudoscalar excitations by adding suitable interactions to the string and show that it is possible to have a least massive pseudoscalar bound state with only bosonic fields. We finally show how certain interactions in the bulk UV theory give rise to background field interactions in the effective string theory.

We present a method to recover and study the projected gravitational tidal forces from a galaxy survey containing little or no redshift information. The method and the physical interpretation of the recovered tidal maps as a tracer of the cosmic web are described in detail. We first apply the method to a simulated galaxy survey and study the accuracy with which the cosmic web can be recovered in the presence of different observational effects, showing that the projected tidal field can be estimated with reasonable precision over large regions of the sky. We then apply our method to the 2MASS survey and present a publicly available full-sky map of the projected tidal forces in the local Universe. As an example of an application of these data we further study the distribution of galaxy luminosities across the different elements of the cosmic web, finding that, while more luminous objects are found preferentially in the most dense environments, there is no further segregation by tidal environment.

We prove the asymptotic roundness under normalized Gauss curvature flow provided entropy is initially small enough.

We report the phase diagram for the superconducting system (${^{7}}$Li${_{1-x}}$Fe${_{x}}$OD)FeSe and contrast it with that of (Li${_{1-x}}$Fe${_{x}}$OH)FeSe both in single crystal and powder forms. Samples were prepared via hydrothermal methods and characterized with laboratory and synchrotron X-ray diffraction, high-resolution neutron powder diffraction (NPD), and high intensity NPD. We find a correlation between the tetragonality of the unit cell parameters and the critical temperature, $T_{c}$, which is indicative of the effects of charge doping on the lattice and formation of iron vacancies in the FeSe layer. We observe no appreciable isotope effect on the maximum $T_{c}$ in substituting H by by D. The NPD measurements definitively rule out an antiferromagnetic ordering in the non-superconducting (Li${_{1-x}}$Fe${_{x}}$OD)FeSe samples below 120 K, which has been reported in non-superconducting (Li${_{1-x}}$Fe${_{x}}$OH)FeSe.$^{1}$ A likely explanation for the observed antiferromagnetic transition in (Li${_{1-x}}$Fe${_{x}}$OH)FeSe samples is the formation of impurities during their preparation such as Fe${_{3}}$O${_{4}}$ and LixFeO2, which express a charge ordering transition known as the Verwey transition near 120 K. The concentration of these oxide impurities is found to be dependent on the concentration of the lithium hydroxide reagent and the use of H${_{2}}$O vs. D${_{2}}$O as the solvent during synthesis. We also describe the reaction conditions that lead to some of our superconducting samples to exhibit ferromagnetism below $T_{c}$.

An analysis is made of the role played by the gas environment in neutron-mirror-neutron and neutron-antineutron oscillations. In the first process the interaction with the ambient medium induces a refraction energy shift which plays the role of an extra magnetic field. In the second process antineutron annihilation in practice might lead to strong decoherence, which should be taken into account in experiments with free neutrons looking for the neutron to antineutron transformation.

False discovery rate (FDR) control has recently proved to be critical in scientific applications involving testing multiple hypotheses on the same dataset, and is intricately related to the recent public controversy regarding reproducibility of scientific results. In many practical applications of multiple hypothesis testing, the hypotheses can be naturally partitioned into groups, and one may not only want to control the number of falsely discoveries (wrongly rejected individual hypotheses), but also the number of falsely discovered groups of hypotheses (where a group is said to be falsely discovered if at least one hypothesis within that group is rejected, when in reality none of the hypotheses within that group should have been rejected). In this paper, we introduce the p-filter, a generalization of the standard FDR procedure by Benjamini and Hochberg [1], and prove that our proposed method can simultaneously control the overall FDR at the finest level (individual hypotheses treated separately) and the group FDR at coarser levels (when such groups are user-specified). We then generalize the p-filter procedure even further to handle multiple partitions of hypotheses, since that might be natural in many applications. For example, in neuroscience experiments, we may have a hypothesis for every (discretized) location in the brain, and at every (discretized) timepoint: does the stimulus correlate with activity in location x at time t after the stimulus was presented? In this setting, one might want to group hypotheses by location or by time (or both). Our procedure naturally generalizes to handle multiple possible partitions of the hypotheses; in the above example, this would amount to controlling overall FDR over all voxels and time points, and FDR at each individual voxel over all time points, and FDR over all voxels at a particular time point.

This paper proposes a boosting-based solution addressing metric learning problems for high-dimensional data. Distance measures have been used as natural measures of (dis)similarity and served as the foundation of various learning methods. The efficiency of distance-based learning methods heavily depends on the chosen distance metric. With increasing dimensionality and complexity of data, however, traditional metric learning methods suffer from poor scalability and the limitation due to linearity as the true signals are usually embedded within a low-dimensional nonlinear subspace. In this paper, we propose a nonlinear sparse metric learning algorithm via boosting. We restructure a global optimization problem into a forward stage-wise learning of weak learners based on a rank-one decomposition of the weight matrix in the Mahalanobis distance metric. A gradient boosting algorithm is devised to obtain a sparse rank-one update of the weight matrix at each step. Nonlinear features are learned by a hierarchical expansion of interactions incorporated within the boosting algorithm. Meanwhile, an early stopping rule is imposed to control the overall complexity of the learned metric. As a result, our approach guarantees three desirable properties of the final metric: positive semi-definiteness, low rank and element-wise sparsity. Numerical experiments show that our learning model compares favorably with the state-of-the-art methods in the current literature of metric learning.

The spread of Ebola virus in 2014 is unprecedented. The epidemic is still affecting West Africa, exacerbated by extraordinary socioeconomic disadvantages and health system inadequacies. With the aim of understanding, predicting, and control the propagation of the virus in the populations of affected countries, it is crucial to model the dynamics of the virus and study several strategies to control it. In this paper, we present a very simple mathematical model that describes quite well the spread of Ebola. Then, we discuss several strategies for the control of the propagation of this lethal virus into populations, in order to predict the impact of vaccine programmes, treatment, and the impact of educational campaigns.

An approach is proposed for evaluating superexchange inter-site interactions in strongly-correlated materials. This approach is based on the single-site dynamical mean-field theory (DMFT) in conjunction with the atomic (Hubbard-I) approximation for the local self-energy. Starting from the local-moment paramagnetic state described by DMFT with Hubbard-I we derive inter-site interactions by considering the response of the DMFT grand potential to small fluctuations of atomic configurations on two neighbouring sites. The result can be subsequently cast into the conventional form of superexchange couplings between local dipolar and multipolar moments. The present method is validated by applying it to one-band and two-band $e_g$ Hubbard models on the simple-cubic 3$d$ lattice.

We study the left-right action of $\operatorname{SL}_n \times \operatorname{SL}_n$ on $m$-tuples of $n \times n$ matrices with entries in an infinite field $K$. We show that invariants of degree $n^2- n$ define the null cone. Consequently, invariants of degree $\leq n^6$ generate the ring of invariants if $\operatorname{char}(K)=0$. We also prove that for $m \gg 0$, invariants of degree at least $n\lfloor \sqrt{n+1}\rfloor$ are required to define the null cone. We generalize our results to matrix invariants of $m$-tuples of $p\times q$ matrices, and to rings of semi-invariants for quivers. For the proofs, we use new techniques such as the regularity lemma by Ivanyos, Qiao and Subrahmanyam, and the concavity property of the tensor blow-ups of matrix spaces. We will discuss several applications to algebraic complexity theory, such as a deterministic polynomial time algorithm for non-commutative rational identity testing, and the existence of small division-free formulas for non-commutative polynomials.

We consider the topology for a class of hypersurfaces with highly nonisolated singularites which arise as exceptional orbit varieties of a special class of prehomogeneous vector spaces, which are representations of linear algebraic groups with open orbits. These hypersurface singularities include both determinantal hypersurfaces and linear free (and free*) divisors. Although these hypersurfaces have highly nonisolated singularities, we determine the topology of their Milnor fibers, complements and links. We do so by using the action of linear algebraic groups beginning with the complement, instead of using Morse type arguments on the Milnor fibers. This includes replacing the local Milnor fiber by a global Milnor fiber which has a complex geometry resulting from a transitive action of an appropriate algebraic group, yielding a compact model submanifold for the homotopy type of the Milnor fiber. The topology includes the (co)homology (in characteristic 0, and 2 torsion in one family) and homotopy groups, and we deduce the triviality of the monodromy transformations on rational (or complex) cohomology. The cohomology of the Milnor fibers and complements are isomorphic as algebras to exterior algebras or for one family, modules over exterior algebras; and cohomology of the link is, as a vector space, a truncated and shifted exterior algebra, for which the cohomology product structure is essentially trivial. We also deduce from Bott's periodicity theorem, the homotopy groups of the Milnor fibers for determinantal hypersurfaces in the stable range as the stable homotopy groups of the associated infinite dimensional symmetric spaces. Applying a Theorem of Oka we obtain a class of formal linear combinations of exceptional orbit hypersurfaces which have Milnor fibers which are homotopy equivalent to joins of the compact model submanifolds.

The naming game has become an archetype for linguistic evolution and mathematical social behavioral analysis. In the model there are $N$ individuals and $K$ words, and we primarily consider arbitrary $K$. In particular, we develop a robust method that handles the case when $K=O(N)$. The initial condition plays a crucial role in the ordering of the system. We find that if the system has high Shannon entropy, then the system has a higher consensus time and a lower critical fraction of zealots compared to low entropy states. We also provide estimates which show that the critical number of committed agents decreases with the number of opinions, and grows with the community size for each word. These results reinforce the maxims "divide and conquer" and "strength in numbers" in opinion propagation.

We present measurements of the diffusion-limited growth of ice crystals from water vapor at different supersaturation levels in air at a temperature of -15 C. Starting with thin, c-axis ice needle crystals, the subsequent growth morphologies ranged from blocky structures on the needle tips (at low supersaturation) to thin faceted plates on the needle tips (at high supersaturation). We successfully modeled the experimental data, reproducing both growth rates and growth morphologies, using a cellular-automata method that yields faceted crystalline structures in diffusion-limited growth. From this quantitative analysis of well-controlled experimental measurements, we were able to extract information about the attachment coefficients governing ice growth under different circumstances. The results strongly support previous work indicating that the attachment coefficient on the prism surface is a function of the width of the prism facet. Including this behavior, we created a comprehensive model at -15 C that explains all the experimental data. To our knowledge, this is the first demonstration of a kinetic model that reproduces a range of diffusion-limited ice growth behaviors as a function of supersaturation.

We numerically study water wave packets on a spatially varying counter-current in the presence of surface tension. Depending on the details of the velocity profile, we show that traversable and bi-directional analogue wormholes exist in fluid mechanics. The limitations on traversability of wormholes in general relativity are absent here because of the dispersion of water waves and the ability to form flow profiles that are not solutions of Einstein's equations. We observe that negative energy can be trapped between analogue horizons forming a LASER-like cavity. Six horizons are involved in the trapping cavity because of the existence of two dispersive scales, in contrast to previous treatments which considered two horizons and one dispersive scale.

Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

We describe a real-time video retrieval framework based on short text input in which weakly labelled training samples from the web are obtained, after the query is known. Concept discovery methods in such a setting train hundreds of detectors at test time and apply them to every frame in the video database. Hence, they are not practical for use in a text based video retrieval setting. We show that an efficient visual representation for a new query can be constructed on-line that enables matching against the test set in real-time. We evaluate a few combinations of encoding, pooling, and matching schemes that are efficient and find that such a system can be built with surprisingly simple and well-known components. We are not only able to construct and apply query models in real-time, but with the help of a re-ranking scheme, we also outperform state-of-the-art methods by a significant margin.

Let $K$ be a finite extension of $\mathbf{Q}_p$. We use the theory of $(\varphi,\Gamma)$-modules in the Lubin-Tate setting to construct some corestriction-compatible families of classes in the cohomology of $V$, for certain representations $V$ of $\mathrm{Gal}(\overline{\mathbf{Q}}_p/K)$. If in addition $V$ is crystalline, we describe these classes explicitly using Bloch-Kato's exponential maps. This allows us to generalize Perrin-Riou's period map to the Lubin-Tate setting.

Low-temperature grid generated turbulence is studied by using numerical simulations of the Gross-Pitaevskii equation. The statistics of regularized velocity increments are studied. Increments of the incompressible velocity are found to be skewed in the inertial range. Results are later confronted with the (quasi) homogeneous and isotropic Taylor-Green flow, revealing the universality of the statistics. For this flow, the statistics are found to be intermittent and a Kolmogorov constant close to the one of classical fluid is found for the second order structure function.

The features of the azimuthal-angle overfocusing of atoms sputtered from the surface of the Ni (001) face are studied by molecular dynamics computer simulation. The signal of overfocused atoms is found to be multi-valued with respect to the initial azimuthal angle due to different mechanisms of scattering. The overfocused atoms form separate maximum and can be detected in experiments with angle and energy resolution separately from the focused and the own atoms.

We examine a mechanism by which excitons undergo ultrafast relaxation in common monolayer transition metal dichalcogenides. It is shown that at densities $\approx$ 1$\times$ 10$^{11}$ cm$^{-2}$ and temperatures $ \le 60$ K, excitons in well known monolayers (MoS$_2$, MoSe$_2$, WS$_2$ and WSe$_2$) exist as point-like structureless electron-hole quasi-particles. We evaluate the average rate of exciton energy relaxation due to acoustic phonons via the deformation potential and the piezoelectric coupling mechanisms and examine the effect of spreading of the excitonic wavefunction into the region perpendicular to the monolayer plane. Our results show that the exciton relaxation rate is enhanced with increase in the exciton temperature, while it is decreased with increase in lattice temperature. Good agreement with available experimental data is obtained when the calculations are extrapolated to room temperatures. A unified approach taking into account deformation potential and piezoelectric coupling mechanisms shows that exciton relaxation induced by phonons is comparable to defect assisted scattering and trapping of excitons by surface states in monolayer transition metal dichalcogenides.

We propose new gauge invariant actions for open NS, heterotic NS, and closed NS-NS superstring field theories. They are based on the large Hilbert space, and have Wess-Zumino-Witten-like expressions which are the $\mathbb{Z}_{2}$-reversed versions of the conventional WZW-like actions. On the basis of the procedure proposed in arXiv:1505.01659, we show that our new WZW-like actions are completely equivalent to $A_{\infty }/L_{\infty }$ actions proposed in arXiv:1403.0940 respectively.

In order to study AS-regular algebras of dimension 5, we consider dimension 5 graded iterated Ore extensions generated in degree one. We classify the possible degrees of relations and structure of the free resolution for extensions with 3 and 4 generators. We show that every known type of algebra of dimension 5 can be realized by an Ore extension and we consider which of these types cannot be realized by an enveloping algebra.

For the asymmetric spherical dimer of an endohedrally confined atom and a host fullerene, an innershell vacancy of either system can decay through the continuum of an outer electron hybridized between the systems. Such decays, viewed as coherent superpositions of the single-center Auger and two-center inter-Coulombic (ICD) amplitudes, are found to govern leading decay mechanisms in noble-gas endofullerenes, and are likely omnipresent in this class of nanomolecules. A comparison between resulting autoionizing resonances calculated in the photoionization of Ar@C60 and Kr@C60 exhibits details of the underlying processes.

We study the partition function of 4d supersymmetric gauge theories with an R-symmetry on Euclidean $S^3\times S_\beta^1$, with $S^3$ the unit-radius squashed three-sphere, and $\beta$ the circumference of the circle. For superconformal theories, this partition function coincides (up to a Casimir energy factor) with the 4d superconformal index. The partition function can be computed exactly using supersymmetric localization of the gauge theory path-integral. It takes the form of an elliptic hypergeometric integral, which may be viewed as a matrix-integral over the moduli space of the holonomies of the gauge fields around $S_\beta^1$. At high temperatures ($\beta\to 0$, corresponding to the hyperbolic limit of the elliptic hypergeometric integral) we obtain from the matrix-integral a quantum effective potential for the holonomies. The effective potential is proportional to the temperature. Therefore the high-temperature limit further localizes the matrix-integral to the locus of the minima of the potential. If the effective potential is positive semi-definite, the leading high-temperature asymptotics of the partition function is given by the formula of Di Pietro and Komargodski, and the subleading asymptotics is connected to the Coulomb branch dynamics on $R^3\times S^1$. In theories where the effective potential is not positive semi-definite, the Di Pietro-Komargodski formula needs to be modified. In particular, this modification occurs in the SU(2) theory of Intriligator-Seiberg-Shenker, and the SO(N) theory of Brodie-Cho-Intriligator, both believed to exhibit "misleading" anomaly matchings, and both believed to yield interacting superconformal field theories with $c<a$. Two new simple tests for dualities between 4d supersymmetric gauge theories emerge as byproducts of our analysis.

In this work, we present a MCTS-based Go-playing program which uses convolutional networks in all parts. Our method performs MCTS in batches, explores the Monte Carlo search tree using Thompson sampling and a convolutional network, and evaluates convnet-based rollouts on the GPU. We achieve strong win rates against open source Go programs and attain competitive results against state of the art convolutional net-based Go-playing programs.

We prove Harnack inequalities for hypersurfaces evolving on the unit sphere either by a 1-homogeneous convex curvature function or by the $p$-power of mean curvature with $0<p<1$.