id
stringlengths 20
32
| text
stringlengths 100
121k
| image_path
stringlengths 15
235
|
|---|---|---|
arxivcap_1912.05084_3 | Title: Bayesian Copula Density Deconvolution for Zero-Inflated Data in Nutritional Epidemiology | Caption: =10pt Graphical model depicting the dependency structure of the generative deconvolution model described in Section <ref> for one episodically consumed component $X_{1}$ and one regularly consumed component $X_{2}$. The unfilled and shaded nodes with solid boundaries signify latent and observable variables, respectively. The filled node with dashed boundary may be observed on some of occasions and latent on others. | 1912.05084_3.jpg |
arxivcap_2005.06146_1 | Title: Neutron spin resonance in a quasi-two-dimensional iron-based superconductor | Caption: {Comparison between the resonant energy $E_R$ and the total superconducting gap $\Delta_{tot}$. } a, Dispersion of the spin resonance above $\Delta_{tot}(Q)$ in KCa$_2$Fe$_4$As$_4$F$_2$. The gaps are summed on two Fermi pockets linked by longitudinally incommensurate wave vectors $Q+d$ and $Q-d$. The horizontal bars are the peak width obtained from the gaussian fittings in Fig. 3c. The shadow areas mark the size of electron pocket at $M$ point <cit.>. b, The ratio of $E_R/\Delta_{tot}$ in 12442-type FeSCs and other compounds. Here we use three values of $\Delta_{tot}$ for KCa$_2$Fe$_4$As$_4$F$_2$ ($\Delta_{tot}=$ 13.2 meV ($\beta_2+\delta$), 10.9 meV ($\alpha+\delta$), 9.4 meV ($\beta_1+\delta$) from ARPES results), and two values for CsCa$_2$Fe$_4$As$_4$F$_2$ ($\Delta_L+\Delta_L$= 15 meV and $\Delta_L+\Delta_S$= 9 meV from $\mu$SR results). | 2005.06146_1.jpg |
arxivcap_1906.05326_10 | Title: Magnetism and superconductivity in Fe$_{1+y}$Te$_{1-x}$Se$_x$ | Caption: Difference in INS for FeTe$_{0.6}$Se$_{0.4}$ ($T_c=14$ K) between superconducting ($T=1.4$ K) and normal ($T=25$ K), plotted as $\chi''$ weighted by the energy transfer. (a) energy dependence along high-symmetry directions; (b) Q dependence for data integrated from 1 to 11 meV. Reprinted with permission from Ref. <cit.>, 2014 by the American Physical Society. | 1906.05326_10.jpg |
arxivcap_2303.06827_5 | Title: Kernel Density Bayesian Inverse Reinforcement Learning | Caption: Evaluating manually curated feature-based reward in the sepsis environment. We plot the EVD for 100 reward samples from KD-BIRL and 100 trajectories from AVRIL. KD-BIRL's EVDs are more concentrated around 0 than those of AVRIL, indicating that the reward samples from KD-BIRL's posterior better replicate the demonstrations from the data-generating reward. | 2303.06827_5.jpg |
arxivcap_1802.01268_19 | Title: ASMCNN: An Efficient Brain Extraction Using Active Shape Model and Convolutional Neural Networks | Caption: 3D Box plots of Dice coefficient, Jaccard index, Average Hausdorff distance, Sensitivity and Specificity for OASIS dataset. | 1802.01268_19.jpg |
arxivcap_2205.02212_6 | Title: Quantum Computing Approaches for Mission Covering Optimization | Caption: QPU Results (a) Average Relative Cost versus the number of qubits used. The relative cost is the mission plus the precedence cost obtained minus the lowest possible mission plus precedence cost that can be achieved (while keeping constraints) for the problem; The lowest possible cost is found by brute force: Relative Cost = Algorithm's Solution Cost - Brute Force Solution's Cost. Negative relative costs imply that the solution violated constraints. (b) Average number of constraints violated versus the number of qubits used. The number of constraints counts up how many extras qubit are on within a column in the matrix view representation. This count includes the amount of columns that do not have any qubits on at all. | 2205.02212_6.jpg |
arxivcap_2108.09666_11 | Title: Relational Embedding for Few-Shot Classification | Caption: Visualization of cross-correlation on ImageNet. (a): Top 10 matches in $\bC$ (initial cross-correlation). (b): Top 10 matches in $\hat{\bC} = h(\bC)$ (refined cross-correlation). Unreliable matches are filtered through $h(\cdot)$. | 2108.09666_11.jpg |
arxivcap_1902.04007_1 | Title: Scaling laws for harmonically trapped two-species mixtures at thermal equilibrium | Caption: Total interaction energy $\bar{E}_{\mathrm{int}}$ versus the number of particles $N$, with both quantities shown on logarithmic scales, for a broad range of $N$ encompassing also the few-body case. Deviations from a power-law scaling at small number of particles are attributable to the few-body dynamics which does not allow for meaningful comparison of the time-averaged total interaction energy to the corresponding one obtained by a canonical ensemble average. Including points from the few-body cases does indeed affect the best fit significantly. A scaling exponent $\alpha=1.80 \pm 0.04$ is obtained for a global fit (red line). Considering only the four rightmost points gives $\alpha=1.61 \pm 0.04$, while the eight rightmost result in $\alpha=1.62 \pm 0.02$, showing robustness of the fit for large $N$. The inset shows the dependence of the standard deviation of the total interaction energy on the number of time steps used for evaluating its average value at the end of the simulation, for the case of $N$=50 particles. The optimal choice is a compromise between the larger standard deviations for smaller time sequences, and the need to avoid bias due to possible residual thermalization dynamics for larger size of the sample. The coupling strength is $\gamma=2.0$, while the parameters $\lambda, \beta_A, \beta_B, m_A, m_B, \omega_A, \omega_B$ have the same values as in Fig. <ref>. The error bars in the inset correspond to one standard deviation from the average value, while the errors on the scaling exponent $\alpha$ here and in the following figures are evaluated as one standard deviation in the least squares analysis. Based on this analysis, we use a minimum of $N$=50 particles for each species, and $10^3$ time steps for the time averaging in all other figures. | 1902.04007_1.jpg |
arxivcap_2305.05287_4 | Title: Nanomechanical Photothermal Near Infrared Spectromicroscopy of Individual Nanorods | Caption: 2D maps of the same region at $\lambda_{exc} = 808$ nm for three different polarization angles $\theta_{pol}$: 90°, 135°, 180°. The two signals are from two individual nanorods. For the perpendicular polarizations, 90° and 180°, the absorbers behave in an opposite way, meaning that they are almost perpendicular one to each other, while absorbing almost the same amount of light for the central map ($\theta_{pol} = 135$°). | 2305.05287_4.jpg |
arxivcap_2303.14786_7 | Title: Detecting Low-Mass Perturbers in Cluster Lenses using Curved Arc Bases | Caption: Posterior probability distributions of the perturber parameters for the two different data sets with different arc magnifications. The green contours correspond to "perturber 1", which is also shown in Figs. <ref> and <ref>. The gray contours correspond to a perturber with the same properties as "perturber 1" but placed on top of an arc whose magnification is 50% less than that of "perturber 1". | 2303.14786_7.jpg |
arxivcap_2303.06763_1 | Title: MigraR: an open-source, R-based application for analysis and quantification of cell migration parameters | Caption: The data sources of MigraR are MSExcel (xlsx, csv) or text files generated by dedicated cell-tracking software; the example provided was deployed by Imaris{TM}. In these files, data is organized in four columns, as follows: column A) $x$ coordinates; column B) $y$ coordinates; column D) time; column E) TrackID (the cell track identifier, attributed by the cell tracking software). | 2303.06763_1.jpg |
arxivcap_1906.05326_11 | Title: Magnetism and superconductivity in Fe$_{1+y}$Te$_{1-x}$Se$_x$ | Caption: Thermal evolution of the magnetic scattering at $\hbar\omega=5$ meV along ${\bf Q}=(1-K,K,0)$ for the Ni02 sample (see text) measured at (a) 100 K, (b) 40 K, (c) 15 K, (d) 2.8 K; (e) related results for the Ni04 sample plotted as an intensity contour map in temperature-wave-vector space. The data have been smoothed. The yellow and black symbols in (e) denote the corresponding peak positions for the Ni02 sample (yellow squares) and for a superconducting Fe$_{1+\delta}$Te$_{0.35}$Se$_{0.65}$ sample <cit.>. Reprinted with permission from Ref. <cit.>, 2012 by the American Physical Society. | 1906.05326_11.jpg |
arxivcap_1912.05084_6 | Title: Bayesian Copula Density Deconvolution for Zero-Inflated Data in Nutritional Epidemiology | Caption: =10pt Results for simulated data sets with sample size $n = 1000$, $q=2$ episodic components and $p=1$ regular components, each subject having $m_{i}=3$ replicates, for the data set corresponding to the 25th percentile $3$-dimensional ISE. The estimated (in blue) probabilities of reporting positive consumption $P_{\ell}(X_{\ell})$ for the two episodic components, estimated by our method. In all panels, the black lines represent the truth. | 1912.05084_6.jpg |
arxivcap_2202.01503_4 | Title: Global sensitivity analysis based on Gaussian-process metamodelling for complex biomechanical problems | Caption: Grown tumour in its microenvironment (A, B, C) with distribution of nanoparticles (D) and necrotic tumour cells (E). A. Volume fraction of living tumour cells $\varepsilon^{\textsf{LTC}}$; B. Volume fraction of vasculature $\varepsilon^v$; C. Pressure $p^l$ in the IF; D. Mass fraction of nanoparticles in the IF $\NPl$; E. Mass fraction of necrotic tumour cells $\NTC$; Subfigures D and E as an example present the result for the mean values of the uncertain input parameters. | 2202.01503_4.jpg |
arxivcap_2202.00203_5 | Title: Constraining low redshift [CII] Emission by Cross-Correlating FIRAS and BOSS Data | Caption: Galaxy counts as a function of redshift for the BOSS LOWZ (blue) and CMASS (red) populations. Plotted in yellow is the expected FIRAS noise angular power spectrum, $C_{\ell}$, at $\ell=0$ for the corresponding frequencies of the fine-structure line. For most of the angular modes accessible to FIRAS, the noise power spectrum will be higher, growing roughly proportional to the inverse of the square of the FIRAS beam and scan window function. For $z<0.2$, the FIRAS noise grows rapidly. The shaded blue and red rectangles indicate the redshift ranges used for the cross-power analysis of the FIRAS data with the LOWZ and CMASS samples. These regions were selected because both the FIRAS thermal noise and galaxy shot noise are low. | 2202.00203_5.jpg |
arxivcap_1802.01428_1 | Title: Re-thinking non-inferiority: a practical trial design for optimising treatment duration | Caption: Prediction curves (red) from 100 simulations against the true data generating curve (black) for all the eight scenarios. | 1802.01428_1.jpg |
arxivcap_2212.13136_8 | Title: Fewer is More: Efficient Object Detection in Large Aerial Images | Caption: Examples of detection results on the driving-scene dataset TT100K (top row) and 4K video dataset Okutama-Action (bottom row). Here, the symbols (top row) denote different traffic-sign classes, whose details can be found in <cit.>. | 2212.13136_8.jpg |
arxivcap_2206.01855_13 | Title: A possible surviving companion of the SN Ia in the Galactic SNR G272.2-3.2 | Caption: $g$ $vs$ $g - r$ magnitudes, at the distance of the SNR G272.2-3.2 (taken here as 2 kpc), of the post-explosion evolutionary tracks of MS (red) and He (magenta) star companions (from Pan et al. 2014a) and location of possible sdB companions (green) (from Meng & Li 2019), compared with our sample of stars (blue filled pentagons) from {Gaia} EDR3 and with the larger sample from the {DECaPS} survey (black crosses), covering the same area of the sky but with no constraints on distance there. The stars have been dereddened as discussed in Section 5. | 2206.01855_13.jpg |
arxivcap_2303.06763_5 | Title: MigraR: an open-source, R-based application for analysis and quantification of cell migration parameters | Caption: Representative graphical displays produced by MigraR based on two time-lapse microscopy videos: i) 'Directional movement', where cells were stimulated to move in the direction of a chemoattractant placed on their left; ii) 'Random movement', where cells were not exposed to any chemoattractant, hence did not exhibit one prevalent directional choice. | 2303.06763_5.jpg |
arxivcap_1806.08867_4 | Title: xGEMs: Generating Examplars to Explain Black-Box Models | Caption: We test whether ResNet models $f^1_{\phi}$ and $f^2_{\phi}$, both trained to detect hair color but on different data distributions are with gender. Two samples for classifiers $f^1_{\phi}$ (first sub row) and $f^2_{\phi}$ (second sub row) are shown. The leftmost image is the original figure, followed by its reconstruction from the encoder $F_{\psi}$. Reconstructions are plotted as Algorithm <ref> (with $\lambda=0.01$) progresses toward crossing the decision boundary. The red bar indicates change in hair color label indicated at the top of each image along with the confidence of prediction. The label at the bottom indicates gender as predicted by $\hat{g}$. For both samples, classifier $f^1_{\phi}$, trained on biased data changes the gender ($1^{st}$ and $3^{rd}$ rows) while crossing the decision boundary whereas the other black-box does not. | 1806.08867_4.jpg |
arxivcap_1111.0239_4 | Title: Correlated phases of bosons in the flat lowest band of the dice lattice | Caption: (Color online) Data for $N=8$ atoms at $\nu=1/2$ on a lattice of $4\times 2 $ unit cells. (a) Two-point correlation function $\langle\hat n_r \hat n_0 \rangle$ of the groundstate $|0\rangle$ for $\vec\theta=(0,\pi)$. The reference site `0' is visually highlighted. Axes carry units $a=|\vec v_1|$, $h=\sqrt{3/4}\,|\vec v_1|$. (b) Density $\langle\hat n_r \rangle$ of the symmetry-broken crystal state $|S\rangle$ (see main text). (c) Fluctuations $\langle\hat n_r^2 \rangle - \langle n_r \rangle^2$ (blue circles), and condensate wavefunction $\vec v_0$ of $|S\rangle$ (red arrows). (d) Scaling of the condensate fraction $\lambda_0/N$ against $N^{-1}$, for superposition states $|S\rangle$ and pinned states with pinning centers of strength $V_p$ on crystal sites. | 1111.0239_4.jpg |
arxivcap_2005.06146_2 | Title: Neutron spin resonance in a quasi-two-dimensional iron-based superconductor | Caption: {Neutron spin resonance and pairing symmetry in cuprates and iron-based high-$T_c$ superconductors.}
a, Single-band $d-$wave pairing in hole-doped cuprates. b, Sign-reversed multi-band $s^{\pm}-$pairing in FeSCs. c, Sign-preserved multi-band $s^{++}-$pairing in FeSCs. {d, e, f}, Dispersion of spin excitations / resonance in the superconducting state corresponding to the pairing symmetries in {a, b, c}. g, Comparison between the sharp resonant peak under $s^{\pm}-$pairing, broad resonant peak under $d-$pairing and broad spin-resonance-like hump under $s^{++}$-wave pairing in FeSCs as shown by the local susceptibility $\chi^{\prime\prime}(Q,\omega)$ below $T_c$. h, Summarized spin resonance energy $E_R$ versus $T_c$ in iron-based superconductors, where the grey dash line $E_R=4.9 k_BT_c$ scales most of them, and $E_R=5.8 k_BT_c$ is usually used in cuprates. Here $\hbar\omega_c$ is the continuum threshold energy in cuprates, and $\Delta_{tot}=|\Delta_k|+|\Delta_{k+Q}|$ is defined as the total superconducting gaps summed on two Fermi surfaces connected by the wavevector $Q$. | 2005.06146_2.jpg |
arxivcap_1802.01428_2 | Title: Re-thinking non-inferiority: a practical trial design for optimising treatment duration | Caption: Prediction curves leading to the largest sABC for each of the eight scenarios with the base-case design, analysing data either with LS3 (blue) or FP (red). | 1802.01428_2.jpg |
arxivcap_2305.05287_6 | Title: Nanomechanical Photothermal Near Infrared Spectromicroscopy of Individual Nanorods | Caption: Simulated field enhancement assuming a laser beam with $\lambda = 840$ nm polarized parallel to the long axis of the nanorod. It takes place in the region surroundings the hemispherical caps of the nanorod, of the order the particle diameter <cit.>. | 2305.05287_6.jpg |
arxivcap_1906.05326_12 | Title: Magnetism and superconductivity in Fe$_{1+y}$Te$_{1-x}$Se$_x$ | Caption: Inelastic magnetic neutron scattering from the SC70 sample (see text) at energy transfers $\hbar\omega=$ 13 meV (a), (b), (c); 10 meV (d), (e), (f); and 7 meV (g), (h), (i). The sample temperatures are 8 K (a), (d), (g); 100 K (b), (e), (h); and 300 K (c), (f), (i). All slices were taken with an energy width of 2 meV. Measurements, covering approximately two quadrants, have been symmetrized to be 4-fold symmetric, consistent with sample symmetry. Intensity scale is the same in all panels, but 13-meV data have been multiplied by 1.5 to improve visibility. Black regions at the center of each panel are outside of the detector range. Panels (j), (k), (l) are model calculations simulating the 7-meV data, using the spin-plaquette model as described in the text, based on weakly correlated slanted UDUD spin plaquettes. The inter-plaquette correlations used in the calculations correspond to (j) 100% stripe, (k) 50% stripe and 50% bicollinear, and (l) 100% bicollinear. Reprinted with permission from Ref. <cit.>, 2016 by the American Physical Society. | 1906.05326_12.jpg |
arxivcap_1109.6687_14 | Title: Extending the Tamari lattice to some compositions of species | Caption: On the left is the Hasse diagram for $n$-tubings of the cycle graph, with the cyclohedron pictured below for comparison. On the right is the Hasse diagram for $n$-tubings of the star graph, with the stellohedron pictured below. | 1109.6687_14.jpg |
arxivcap_1912.05084_7 | Title: Bayesian Copula Density Deconvolution for Zero-Inflated Data in Nutritional Epidemiology | Caption: =10pt Results for simulated data sets with sample size $n = 1000$, $q=2$ episodic components and $p=1$ regular components, each subject having $m_{i}=3$ replicates, for the data sets corresponding to the 25th percentile $3$-dimensional ISEs. The estimated distributions of the two episodic components, normalized by the regular component, estimated by our method (in blue) and by the method of <cit.> (in red). In all panels, the black lines represent the truth. | 1912.05084_7.jpg |
arxivcap_2303.06847_1 | Title: Label Distribution Learning from Logical Label | Caption: An example of using label distribution to describe a natural scene image in (a). The histogram in (b) denotes the logical label, indicating whether a label can describe the image in (a), and the line chart in (c) shows the label distribution, revealing to what degree a label can describe the image in (a). | 2303.06847_1.jpg |
arxivcap_2305.05287_7 | Title: Nanomechanical Photothermal Near Infrared Spectromicroscopy of Individual Nanorods | Caption: Analytical model of the damping mechanisms behind the broadening of the nanorod LSPR resonance. $\Gamma_{bulk}$ represents the mean bulk-like electron scattering rate (grey dashed line); $\Gamma_{surf}$ represents the electron-surface scattering rate (red solid line); $\Gamma_{rad}$ represents the radiative damping (blue solid line). The black solid line indicates the overall damping $\Gamma_{tot}$, sum of the three contributions. Electron-bulk and surface scattering are the major source of plasmonic damping, with the radiative one being almost two orders of magnitude smaller. | 2305.05287_7.jpg |
arxivcap_2210.09493_3 | Title: Strong lensing constraints on primordial black holes as a dark matter candidate | Caption: Constraints from disruption of wide binaries (WB) <cit.>, Eridanus II star cluster surviving possible destruction by dynamical heating (ES) <cit.>, halo dynamical friction (DF) <cit.>, large-scale structure (LSS) (, ), X-ray background from accretion (XB) <cit.>, and our constraint from strong lensing flux ratio analysis (SL). | 2210.09493_3.jpg |
arxivcap_1802.01428_3 | Title: Re-thinking non-inferiority: a practical trial design for optimising treatment duration | Caption: Boxplots comparing results of trial simulations from the eight scenarios varying either (i) the flexible regression method used (LS3, LS5, LSNE, MARS, FP), with total sample size of 504 patients (panel (a)), or (ii) the total sample size between 250 and 1000 patients, using FP (panel (c)). Patients are divided in 7 equidistant duration arms. The red horizontal line indicates $5\%$ sABC. In (b) and (d) we compare $95^{th}$ percentiles from the eight scenarios. | 1802.01428_3.jpg |
arxivcap_2102.12911_15 | Title: Blocks World Revisited: The Effect of Self-Occlusion on Classification by Convolutional Neural Networks | Caption: Illustration of the self-occlusion distribution for $L_1$ and $L_2$ (top), as well as the distributional relation between viewpoint mapping and self-occlusion for $L_1$ and $L_2$ (bottom). | 2102.12911_15.jpg |
arxivcap_2303.14786_8 | Title: Detecting Low-Mass Perturbers in Cluster Lenses using Curved Arc Bases | Caption: Posterior probability distributions of the perturber parameters for a model with a cored NFW profile. The core radius $c$ modifies the density profile as described by Eq. (<ref>). | 2303.14786_8.jpg |
arxivcap_2202.13734_4 | Title: Missing Value Estimation using Clustering and Deep Learning within Multiple Imputation Framework | Caption: MICE imputed data have shown to yield superior and robust classification accuracy compared to single imputation methods, e.g., median and K-nearest neighbor. | 2202.13734_4.jpg |
arxivcap_2212.13218_12 | Title: Multisensor Data Fusion for Reliable Obstacle Avoidance | Caption: Trajectory of the robot before and after fusion map during navigation in a multi-obstacle environment in a static environment | 2212.13218_12.jpg |
arxivcap_2202.00203_6 | Title: Constraining low redshift [CII] Emission by Cross-Correlating FIRAS and BOSS Data | Caption: Mollweide projection of the CMASS over-density map and selection function at $z\sim0.52$, near the redshift peak of CMASS, in Galactic coordinates. There are several points in the over-density map well above 4, but we choose to saturate the scale at 4 to show the broad clustering features. | 2202.00203_6.jpg |
arxivcap_1902.04007_2 | Title: Scaling laws for harmonically trapped two-species mixtures at thermal equilibrium | Caption: Total interaction energy at thermal equilibrium per unit of coupling strength $\gamma$ versus the coupling strength itself for baths made of $10^2$ particles each, and the same temperatures and interaction range as in Fig.<ref>. The plots show evidence of the saturation of the interaction energy in both extremes of strong attractive and repulsive couplings in all dimensions. | 1902.04007_2.jpg |
arxivcap_2205.02281_1 | Title: Three-Body Problem in Modified Dynamics | Caption: The boundaries of stable and unstable systems in $L_{4} $ and $L_{5}$ Lagrange points, defined by $\mu$, for different values of $\epsilon$. Systems within the bullet shape curve are unstable. When $\epsilon$ increases, i.e. when the term proportional to $a_{0}$ begins to dominate the equation of motion, systems with larger mass ratios could be stable in their $L_{4} $ and $L_{5}$ Lagrange points even when $\mu > 0.04$. For example, in the case of $\epsilon \approx 0.5$, i.e. the vertical line in the above plot, systems with a mass ratio of $\mu \lesssim 0.2 $ could still be stable. | 2205.02281_1.jpg |
arxivcap_1104.3706_2 | Title: Detection of a large massive circumstellar disk around a high-mass young stellar object in the Carina Nebula | Caption: Negative grayscale representation of the $8\,\mu$m Spitzer image of the cloud in which the disk object is embedded, with overplotted contours of the $870\,\mu$m LABOCA map. The field-of-view is $\approx 4.5' \times 3.8'$; north is up and east to the left. The sub-mm contour levels increase from 0.06 Jy/beam to 1.5 Jy/beam in equal steps. The position of the disk object is marked by the white circle. | 1104.3706_2.jpg |
arxivcap_2202.01503_5 | Title: Global sensitivity analysis based on Gaussian-process metamodelling for complex biomechanical problems | Caption: Nash-Sutcliffe efficiency $Q^2$ for different training sample set sizes with a tensorised, squared, exponential covariance function and a tensorised Matérn covariance function. $N$ training samples were randomly generated for the main plot. For the detail plot, we repeated the process five times with different training sample sets of sizes $N = [10, 15, 20, 25, 30]$. For reference, the dashed lines in the detail plot are identical to the dashed lines in the main plot. | 2202.01503_5.jpg |
arxivcap_1912.05084_8 | Title: Bayesian Copula Density Deconvolution for Zero-Inflated Data in Nutritional Epidemiology | Caption: =10pt Results for the EATS data sets with sample size $n = 965$, $q=2$ episodic components, milk and whole grains, and $p=2$ regular components, sodium and energy, each subject having $m_{i}=4$ replicates. From top to bottom, the left panels show the estimated densities $f_{X,\ell}(X_{\ell})$ of milk and whole grains, sodium, and energy, respectively, obtained by our method (in blue) and the method of <cit.> (in red). The right panels show the associated distributions of the scaled errors $f_{\epsilon,q+\ell}(\epsilon_{q+\ell})$ and the associated variance functions $v_{\ell}(\wt{X}_{\ell}) = s_{\ell}^{2}(\wt{X}_{\ell})$, estimated by our method. | 1912.05084_8.jpg |
arxivcap_1109.6687_16 | Title: Extending the Tamari lattice to some compositions of species | Caption: The multiplihedron lattice $\Fm_4$ showing the three subintervals that yield primitives via Möbius transformation. | 1109.6687_16.jpg |
arxivcap_1906.05326_13 | Title: Magnetism and superconductivity in Fe$_{1+y}$Te$_{1-x}$Se$_x$ | Caption: Inelastic neutron scattering intensities in the $(H0L)$ plane measured at energy transfer $\hbar\omega= 4$ meV on the SC40 and Ni10 samples (see text). The intensities are scaled by the sample mass for better comparison. Left column are 2D intensity slices, and right column are linear intensity cuts along $(0.5,0,L)$. The q-width of the linear cuts is 0.05 r.l.u. along [100] direction. The panels are (a) and (b): SC40 at 5 K; (c) and (d) SC40 at 300 K; (e) and (f) Ni10 at 5 K; and (g) and (h) Ni10 at 300 K. The white line in the left panels at $H = 0.5$ shows where the $L$ cuts in the right column were taken. The dashed lines in right panels are estimated background values obtained from fitting around $(0.65,0,0)$. The blue solid lines in (d), (f), and (h) are the magnetic form factor for an Fe$^{2+}$ ion scaled to the intensity maximum, and the red solid lines are fits to the data using two symmetric Lorentzians peaked at $L=\pm0.5$. All slices were taken with an energy width of 2 meV. The error bars represent statistical error. Reprinted with permission from Ref. <cit.>, 2017 by the American Physical Society. | 1906.05326_13.jpg |
arxivcap_2305.05287_8 | Title: Nanomechanical Photothermal Near Infrared Spectromicroscopy of Individual Nanorods | Caption: (a) Absorption spectra of the LSPR resonance calculated with the T-matrix method together with effective medium approximation (EMA), for different value of the parameter A, for two different aspect ratios $a$ (3.5, dashed lines; 4, solid lines). (b) LSPR linewidth's dependence on the parameter A for: the analytical model (eq. (<ref>)), empty dark violet downward triangles; FEM, empty black circles; T-matrix plus EMA, empty blue squares; mean value and standard deviation of the measured linewidths, red dot. Inset: linewidth as a function of the LSPR energy for the T-matrix calculation (black solid line) and measurements (empty red dots). | 2305.05287_8.jpg |
arxivcap_1802.01428_4 | Title: Re-thinking non-inferiority: a practical trial design for optimising treatment duration | Caption: Boxplots comparing results of trial simulations from the eight scenarios either varying the number of equidistant arms (panel (a)) between 3 and 20, using FP, or using different designs, ED or NED, comparing four different regression methods (panel (c)). The total sample size is always 504 patients. The red horizontal line indicates $5\%$ sABC. In (b) and (d) we compare $95^{th}$ percentiles from the eight scenarios. In panel (d), there is only one point for NED-LS3, since only in one scenario the $95^{th}$ percentile for sABC was smaller than 0.25. | 1802.01428_4.jpg |
arxivcap_1806.08867_7 | Title: xGEMs: Generating Examplars to Explain Black-Box Models | Caption: Reliability Diagram for Calibration stratified by (potentially protected) attributes of interest (gender): A perfectly calibrated classifier should manifest an identity function. Deviation from the identity function suggests mis-calibration and can be used for model comparison when accuracy and other metrics are comparable. | 1806.08867_7.jpg |
arxivcap_2210.09493_4 | Title: Strong lensing constraints on primordial black holes as a dark matter candidate | Caption: Visualization of rendering area relative to image position. The circles around the image positions represent the rendering area for lensing substructure. On the left, for lens RX J0911+0551, these areas do not overlap for the chosen radius of 0.24". On the right, there is significant overlap between the rendering areas for two images, so a new aperture is drawn around both images to avoid double-placement of PBH in the overlap region. | 2210.09493_4.jpg |
arxivcap_2303.06740_1 | Title: Fine-tuning Strategies for Faster Inference using Speech Self-Supervised Models: A Comparative Study | Caption: WER and inference metrics with or without language modelling for the presented techniques fine-tuned on LibriSpeech-100h. The best techniques, characterized by both low Word Error Rates (WERs) and inference times, are Factor2 and Factor3 downsamplings, located in the bottom left of the figures. The full model is indicated by a blue diamond, while DistilHubert baselines are represented by orange squares. Inference time measurements are shown as a proportion of the measure done with the full model. | 2303.06740_1.jpg |
arxivcap_2005.06146_3 | Title: Neutron spin resonance in a quasi-two-dimensional iron-based superconductor | Caption: {Crystal structure, Fermi surface and the spin resonance mode of KCa$_2$Fe$_4$As$_4$F$_2$} a, Crystal structure with interlaced stacks of CaFeAsF and KFe$_2$As$_2$, where the Fe$_2$As$_2$ bilayers are separated by insulating Ca$_2$F$_2$ blocks. b and c, DFT calculation results on the Fermi surfaces<cit.>. There are 4 visible hole pockets with distinct sizes around $\Gamma$ point ($\alpha, \beta, \gamma_1, \gamma_2$) and 3 small electron pockets ($\delta_{1,2,3}$) around $M$ point. The Fermi pockets corresponding to are shown in the 2D Brillouin zone of 2-Fe unit cell, and each of them consists of different orbitals of Fe$^{2+}$ such as $d_{xy}$, $d_{xz}$, $d_{yz}$, $d_{x^2-y^2}$ and $d_{z^2}$ . d, Sketch picture of the spin resonance mode and its downward dispersion. The intensity are obtained from the neutron counts by subtracting the 40 K data at normal state from 8 K data in the superconducting state. | 2005.06146_3.jpg |
arxivcap_2202.13734_5 | Title: Missing Value Estimation using Clustering and Deep Learning within Multiple Imputation Framework | Caption: The number of monotone patterns versus the percentage of missing values for MAR, MCAR and MNAR type of missingness for default of credit card clients data set. | 2202.13734_5.jpg |
arxivcap_2108.09652_2 | Title: Stable and scalable multistage terahertz-driven particle accelerator | Caption: Stable and staged THz-driven interaction. (a) Consecutive measurement of beam energy distribution with both THz pulses off (the first 50 shots), with one-stage acceleration (the middle 50 shots), and with two-stage acceleration (the final 50 shots); (b) Statistical results of the beam energy distribution for the data in (a). | 2108.09652_2.jpg |
arxivcap_2102.12792_2 | Title: Mixed Variable Bayesian Optimization with Frequency Modulated Kernels | Caption: SVM(left), XGBoost(right) (Mean$\pm$Std.Err. of 5 runs) | 2102.12792_2.jpg |
arxivcap_2210.09493_5 | Title: Strong lensing constraints on primordial black holes as a dark matter candidate | Caption: Effective multiplane convergence, a two-dimensional representation of a full population of line of sight haloes and subhaloes, for a dark matter realization in CDM (left) and with PBH substructure (right). Red corresponds to a density higher to that of the mean dark matter density, while blue corresponds to an underdensity. Black circles are plotted at each of the four quad image positions, and the black curves are the critical curves, which follows the region of maximum image magnification. Small-scale features in the convergence map that appear to track towards the origin are associated with black holes rendered around the path followed by the lensed light rays. Deformation of the critical curve by the PBH population suggests they will strongly perturb image flux ratios. | 2210.09493_5.jpg |
arxivcap_2205.02281_2 | Title: Three-Body Problem in Modified Dynamics | Caption: Aperiodic bounded orbits according to Newtonian and MOD models. Upper panels: The distance of the test particle ( in units of $\mathcal{D}$ ) from the central mass is plotted as a function of time ( in units of $\mathcal{T}$ ) within Newtonian model as well as MOD1 and MOD2. Bottom: The corresponding orbits are shown here ( The coordinates $x$ and $y$ are in units of $\mathcal{D}$. ). | 2205.02281_2.jpg |
arxivcap_1111.0430_1 | Title: Fermi-LAT spectral analysis of Fermi, Planck, Swift and radio selected samples of AGN | Caption: SED of PKS 1124-186 from the soft X-ray sample <cit.>. Red filled points: simultaneous multi-frequency data; green points: $\gamma$-ray data integrated over a period of 2 months centered on the Planck observing period, or ground-based data taken quasi-simultaneously; blue points: $Fermi$-LAT data integrated over 27 months; light gray points: literature or archival data. | 1111.0430_1.jpg |
arxivcap_2102.12804_1 | Title: Compact Groups of Galaxies in Sloan Digital Sky Survey and LAMOST Spectral Survey. II. Dynamical Properties of Isolated and Embedded Groups | Caption: The $^{0.1}r$-band luminosity functions of isolated (red), embedded (blue) CGs, and noncompact groups (black) derived via both a nonparametric (stepwise) and parametric maximum likelihood estimator with ($\alpha$, $M_{\star}$) quoted inside the figure. The dashed line represents the LFs of all the galaxies in Y07 catalog given by <cit.>. All of the LFs are normalized to 1.0 at $^{0.1}M_{r} - 5\log{h} = -20.5$ mag for comparison. | 2102.12804_1.jpg |
arxivcap_2206.01855_14 | Title: A possible surviving companion of the SN Ia in the Galactic SNR G272.2-3.2 | Caption: The HR diagram of the stars of our sample, compared with the theoretical evolutionary paths of Pan et al. (2014a) for main-sequence star companions of SNe Ia after the explosion. The evolutionary tracks cover from the time the SN Ia companions recover hydrostatic equilibrium after being impacted by the SN ejecta to 9,000 years later. The 100, 500, 3,000 and 9,000 yr post-explosion stages are marked by filled squares, stars, triangles and circles, respectively. Star M5-G272 is marked by a magenta dot. The stars have been dereddened as discussed in Section 5. | 2206.01855_14.jpg |
arxivcap_1902.04007_3 | Title: Scaling laws for harmonically trapped two-species mixtures at thermal equilibrium | Caption: Scaling exponent $\alpha$ versus the coupling strength $\gamma$ for the same temperatures and interaction range as in Fig. 4, in the 1D case. A narrow region at small values of $\gamma$ is visible in which anomalous scaling occurs. Notice that the error bars in the region of small and negative $\gamma$ are large enough to make the values compatible with $\alpha=2$ within three standard deviations at most, while the case of anomalous scaling is statistically much stronger for positive values of $\gamma$. Specifically, for our data, at $\gamma=1$ we get $\alpha=1.58 \pm 0.06$ which is about $1.3$ standard deviations from the theoretically expected value $5/3$ discussed in the analytical section. The analysis has been repeated for the case of 2D and 3D systems at different coupling strenghts and all the other parameters kept constant as in the 1D case, obtaining exponents of $\alpha=1.85 \pm 0.02 ~(\gamma=1)$, $\alpha=1.74 \pm 0.05 ~(\gamma=2)$, $\alpha=1.82 \pm 0.04 ~(\gamma=20)$, for the 2D case, and $\alpha=2.03 \pm 0.04 ~(\gamma=1)$, $\alpha=2.03 \pm 0.04 ~(\gamma=2)$, $\alpha=2.26 \pm 0.12 ~(\gamma=20)$, for the 3D case. A comprehensive analysis of anomalous scaling for the higher-dimensionality cases will be the subject of future investigation, including the case of anisotropic trapping. | 1902.04007_3.jpg |
arxivcap_2202.00203_7 | Title: Constraining low redshift [CII] Emission by Cross-Correlating FIRAS and BOSS Data | Caption: The approximate expected variance in the cross-power diagonal at $z {\sim} 0.4$, according to the Gaussian error formula $\langle \Delta C^{x}_\ell(z,z') \Delta C^{x}_\ell(z,z') \rangle \approx (f_{\rm sky}(2\ell + 1))^{-1} C^{\rm IM}_\ell(z,z')C_{\ell}^g(z,z')$. Although the magnitude varies somewhat with redshift, the shape is representative of all redshifts studied. The small number of modes and large galactic foregrounds drive high variance at low $\ell$. For $\ell>20$, the foregrounds have mostly subsided, and thermal noise dominates. For $\ell > 30$, the increase in the noise caused by beam and scan convolution starts to overtake the advantage of extra modes at higher $\ell$. Since the FIRAS data was originally mapped at $N_{\rm side}=16$, spherical harmonics below $\ell=3N_{\rm side}=48$ form a complete basis, and no further information can be extracted by considering higher $\ell$. Dashed lines indicate the bounds of the 5 $\ell$-bins, and the shaded regions show the three bins used in the cross-power analysis. | 2202.00203_7.jpg |
arxivcap_1802.01492_4 | Title: Analysing the Degree of Meshing in Medium Voltage Target Grids - An Automated Technical and Economical Impact Assessment | Caption: Resupply switching sequence for a line fault in an open ring: a. normal operation, b. line fault with protection trip, c. fault isolation, d. resupply | 1802.01492_4.jpg |
arxivcap_2212.13126_1 | Title: Eliminating temporal correlation in quantum-dot entangled photon source by quantum interference | Caption: (color online) Experimental results. When X and XX interference occur simultaneously, (a) Population of the four-photon GHZ state; (b) Coherence of the four-photon GHZ state. (c) Comparison of coherence, population and fidelity for two PBS fusion operations and a single PBS fusion operation. When only one PBS is used for concatenating entangled photon pairs, the coherence is $0.362\pm0.011$ and $0.446\pm0.019$ for X and XX interference, respectively. When two PBSs are applied, the coherence is improved to $0.552\pm0.020$, which exceeds the single PBS situation by more than 3.83 standard deviations and 8.28 standard deviations for XX and X interference, respectively. | 2212.13126_1.jpg |
arxivcap_2202.01503_6 | Title: Global sensitivity analysis based on Gaussian-process metamodelling for complex biomechanical problems | Caption: Gaussian process for $N = 20$ with a tensorised, squared, exponential covariance function. Projected mean $m_{\textsf{GP}i}(X_i)$, projected 95% confidence interval (CI) and training samples $\datatrain$ for the mean of the necrotic fraction of tumour cells $\bar{\omega}^{N\bar{t}}$ (the y-axis labels apply to both figures). | 2202.01503_6.jpg |
arxivcap_2303.14786_9 | Title: Detecting Low-Mass Perturbers in Cluster Lenses using Curved Arc Bases | Caption: Posterior probability distributions of the perturber parameters for a model with a elliptical NFW profile. | 2303.14786_9.jpg |
arxivcap_1401.8053_11 | Title: Hallucinating optimal high-dimensional subspaces | Caption: Bilinear projection model - the inferred most similar modes of variation contained within two subspaces representing face appearance variation of the same person in different illumination conditions and at different training scales. In each subfigure, which corresponds to a different training-query scale discrepancy, the top pair of images represents appearance extracted by the naïve algorithm of Section <ref> (as the left-singular and right-singular vectors of ${\mathbf{B}_Y}^T~\mathbf{B}^*_X$); the bottom pair is extracted by the proposed method (as the left-singular and right-singular vectors of ${\mathbf{B}_Y}^T~\mathbf{B}_{Xc}$). | 1401.8053_11.jpg |
arxivcap_1806.08744_11 | Title: Control Plane Compression | Caption: Abstractions for a network running BGP on a fattree topology using different policies. | 1806.08744_11.jpg |
arxivcap_2303.06740_2 | Title: Fine-tuning Strategies for Faster Inference using Speech Self-Supervised Models: A Comparative Study | Caption: WER with LM decoding and MACs for the considered methods on WSJ, Buckeye and LibriSpeech-10h sets. While WSJ exhibits results similar to LibriSpeech,reducing the quantity of fine-tuning data causes significant performance drops for the downsampling methods. | 2303.06740_2.jpg |
arxivcap_2108.09637_1 | Title: Graph-Convolutional Deep Learning to Identify Optimized Molecular Configurations | Caption: The change of the PBE0 molecular energy of propane-1-sulfonamide (C3H9NO2S) with respect to $\angle(C1-C2-C3)$ and $\angle(C3-S-N)$ angles is shown in the 3D plot. The most distorted non-equilibrium configuration (top) and the optimized equilibrium structure (bottom) are shown for comparison. The PBE0 atomic forces for the optimized and most distorted configurations are listed as a table. The total force magnitude is given for each case at the bottom of the table to show the substantially suppressed total force magnitude on the atoms for the optimized structure compared to the most distorted configuration. All the presented data are extracted from QM7-X data set <cit.>. | 2108.09637_1.jpg |
arxivcap_2305.05287_9 | Title: Nanomechanical Photothermal Near Infrared Spectromicroscopy of Individual Nanorods | Caption: FEM simulated intensity distribution along a 1D cut line passing in the center of the physical domain at a wavelength of 800 nm, for different silicon nitride slab thicknesses. The intensities are the results to the FEM first step, where only the slab is simulated, without any gold nanorod on top of it. The vertical lines and the relative colors show the positions of each element along the cut line: white, air; grey, silica; yellow, gold; green of different intensities, the different silicon nitride slabs. | 2305.05287_9.jpg |
arxivcap_1109.6638_1 | Title: The Statistical Inefficiency of Sparse Coding for Images (or, One Gabor to Rule them All) | Caption: In panel (a), the factored coding model with 500 basis vectors reconstructs test images (b) using only the largest 64 coefficients. Individual Gabor filters are still visible as they are combined together to form the reconstruction. Panel (c) shows the mean reconstruction error using only the largest $K$ coefficients for each of the first 1000 test images from the CIFAR-10 dataset. For a variety of dictionary sizes ($m$) and code lengths, the factored sparse coding models (solid lines) consistently require about half as many non-zero coefficients in order to produce the same reconstruction error as non-factored models (dashed lines). Estimation error on the RMSE values is negligeable, error bars have been omitted for clarity. (Lower is better.) | 1109.6638_1.jpg |
arxivcap_2202.13734_6 | Title: Missing Value Estimation using Clustering and Deep Learning within Multiple Imputation Framework | Caption: Normalized root mean squared error (NRMSE) comparison between MICE (linear regression-MICE or LR-MICE) and non-MICE algorithms (iterative SVD, matrix factorization, K-nearest neighbor (KNN)). The test data are missing at random (MAR). | 2202.13734_6.jpg |
arxivcap_1802.01492_7 | Title: Analysing the Degree of Meshing in Medium Voltage Target Grids - An Automated Technical and Economical Impact Assessment | Caption: Overview of target grid optimization methodology: the current grid structure (left) is dismantled with and without switching station with assumptions for the grid state in 2030 (middle) and then target grids for switching station, open ring and closed ring grid concept is optimized (right) | 1802.01492_7.jpg |
arxivcap_1104.3706_3 | Title: Detection of a large massive circumstellar disk around a high-mass young stellar object in the Carina Nebula | Caption: $H$-band HAWK-I image of the area around the disk object. The small (green) circles with $2''$ diameter mark the positions of the objects listed in the 2MASS point source catalog. The $5''$ diameter (red) circles mark the locations of the two $10.4\,\mu$m sources detected by <cit.>, and the $18''$ diameter (red) circle marks the peak of the sub-mm emission as seen in our LABOCA map. A grid of J2000 coordinates is shown. | 1104.3706_3.jpg |
arxivcap_1210.8221_5 | Title: Probing Dark Energy Anisotropy | Caption: As Fig. <ref> but for the case of three patches in orthogonal sky directions. Note the change in scale. Now the equation of state estimations are strongly constrained and much less degenerate. | 1210.8221_5.jpg |
arxivcap_1912.05084_9 | Title: Bayesian Copula Density Deconvolution for Zero-Inflated Data in Nutritional Epidemiology | Caption: =10pt Results for the EATS data sets with sample size $n = 965$, $q=2$ episodic components, milk and whole grains, and $p=2$ regular components, sodium and energy, each subject having $m_{i}=4$ replicates. The estimated probabilities of reporting positive consumption $P_{\ell}(X_{\ell})$ for the episodic components milk (left panel) and whole grains (right panel), estimated by our method. | 1912.05084_9.jpg |
arxivcap_2005.06146_5 | Title: Neutron spin resonance in a quasi-two-dimensional iron-based superconductor | Caption: {Photos and X-ray diffraction patterns of KCa$_2$Fe$_4$As$_4$F$_2$ single crystals.} a, Co-aligned crystals on aluminum plates for neutron scattering experiments. b, Sample mount in the [$H, H, L$] scattering plane. c, X-ray diffraction patterns with incident beam along $c-$axis. | 2005.06146_5.jpg |
arxivcap_1111.0430_2 | Title: Fermi-LAT spectral analysis of Fermi, Planck, Swift and radio selected samples of AGN | Caption: SED of PKS B1830-210 from the hard X-ray sample <cit.>. Red filled points: simultaneous multi-frequency data; green points: $\gamma$-ray data integrated over a period of 2 months centered on the Planck observing period, or ground-based data taken quasi-simultaneously; blue points: $Fermi$-LAT data integrated over 27 months; light gray points: literature or archival data. | 1111.0430_2.jpg |
arxivcap_2303.06734_2 | Title: Environmental variability and network structure determine the optimal plasticity mechanisms in embodied agents | Caption: An outline of the network controlling the foraging agent. The sensor layer receives inputs at each time step (the ingredients of the nearest food), which are processed by the plastic layer in the same way as the static sensory network, <ref>. The output of that network is given as input to the motor network, along with the distance $d$ and angle $\alpha$ to the nearest food, the current velocity $v$, and energy $E$ of the agent. These signals are processed through two hidden layers to the final output of motor commands as the linear and angular acceleration of the agent | 2303.06734_2.jpg |
arxivcap_2206.01855_15 | Title: A possible surviving companion of the SN Ia in the Galactic SNR G272.2-3.2 | Caption: Left panel: distribution of the total proper motions of the stars within the 1$^{o}$ radius around the centroid of the SNR G272.2-3.2. Right panel: same, for the tangential velocities. We see a clear outlier, {Gaia} EDR3 5323871314998012928, moving at $\simeq$ 800 (with important error bars due to the parallax uncertainty). A logarithmic scale has been adopted in the vertical axis, for clarity. | 2206.01855_15.jpg |
arxivcap_1806.08852_3 | Title: Multi-Task Handwritten Document Layout Analysis | Caption: F1 measure for OHG using different number of training pages. Task-1 defined as baseline detection only, and Task-1 & 2 as baseline detection plus zone segmentation and labeling. | 1806.08852_3.jpg |
arxivcap_2305.05311_1 | Title: Structured Sentiment Analysis as Transition-based Dependency Parsing | Caption: A sentiment graph (a) is encoded into bi-lexical dependency graphs following the head-first (b) and head-final (c) strategies. | 2305.05311_1.jpg |
arxivcap_2212.13126_2 | Title: Eliminating temporal correlation in quantum-dot entangled photon source by quantum interference | Caption: (color online) Schematic of the protocol extended to multi-photon GHZ generation. (a) Two optical switches guide a pair of interfered photons to the input port of the PBS, then, these photons interfere with the next emission from the QD. Finally, multi-photon GHZ state can be addressed in the time domain. M denotes measurement setup, including single photon detector, TDC, suitable data processing, etc. (b) Signals of optical switches. In the first period, two optical switches can be opened to guide the first pair of entangled photon to pass through the fiber loop. Then, the multi-photon GHZ state can be generated by sequential quantum interference in PBS. | 2212.13126_2.jpg |
arxivcap_1912.05084_10 | Title: Bayesian Copula Density Deconvolution for Zero-Inflated Data in Nutritional Epidemiology | Caption: =10pt Results for the EATS data sets with sample size $n = 965$, $q=2$ episodic components, milk and whole grains, and $p=2$ regular components, sodium and energy, each subject having $m_{i}=4$ replicates. From left to right, the estimated distributions of normalized intakes of milk, whole grains and sodium, normalized by total energy, estimated by our method (in blue) and by the method of <cit.> (in red). | 1912.05084_10.jpg |
arxivcap_2303.14786_10 | Title: Detecting Low-Mass Perturbers in Cluster Lenses using Curved Arc Bases | Caption: Posterior probability distributions of the perturber parameters for a model with a freely varying perturber redshift. The green contours correspond to the foreground line-of-sight perturber with a true redshift of $0.25$. The red contours correspond to the subhalo perturber at the lens redshift of $0.3877$. The blue contours correspond to the background line-of-sight perturber with a true redshift of $0.55$. The true values are shown as dashed lines, and the only one that varies between the three different data sets is the redshift. | 2303.14786_10.jpg |
arxivcap_1902.04007_4 | Title: Scaling laws for harmonically trapped two-species mixtures at thermal equilibrium | Caption: Form factor $F_D$ versus the inverse temperature $\beta$ normalized to the interaction range inverse temperature $\beta_\lambda$. | 1902.04007_4.jpg |
arxivcap_2102.12804_2 | Title: Compact Groups of Galaxies in Sloan Digital Sky Survey and LAMOST Spectral Survey. II. Dynamical Properties of Isolated and Embedded Groups | Caption: Group velocity dispersion ($\sigma_{\text{LOS}}$) as a function of group luminosity ($L_{19.5}$) for isolated CGs (red), embedded CGs (blue), and control noncompact groups (black) on a logarithmic scale with bin size of $0.4$ dex. The open circles, hexagons, and diamonds show the median of $\sigma_{\text{LOS}}$ in each $L_{19.5}$ bin, whereas the 16$^\text{th}$ and 84$^\text{th}$ percentiles are covered by shaded areas. Only the data bins with at least 10 groups are plotted. The green line represents the same scale relation for embedded CGs but their $L_{19.5}$ are replaced by $L_{19.5}^{\text{par}}$. The vertical error bars show the errors of the median $\log\left(\sigma_{\text{LOS}}\right)$ and the horizontal error bars indicate the median absolute deviation of $\log\left(L_{19.5}\right)$ in each bin | 2102.12804_2.jpg |
arxivcap_2210.09493_6 | Title: Strong lensing constraints on primordial black holes as a dark matter candidate | Caption: Posterior distributions created from simulated data using image positions and lensing priors of B1422+231. The posteriors are drawn from the 250 closest samples to the simulated "truth" flux ratios represented by dashed blue lines and corresponding to $M_{\mathrm{PBH}} = 10^{4.1}$M$_\odot$, $f_{\mathrm{PBH}} = 0.02$ on the left and $M_{\mathrm{PBH}} = 10^{5.9}$M$_\odot$, $f_{\mathrm{PBH}} = 0.48$ on the right. | 2210.09493_6.jpg |
arxivcap_2108.09637_2 | Title: Graph-Convolutional Deep Learning to Identify Optimized Molecular Configurations | Caption: Schematic illustration of testing the two trained GC networks on optimized and displaced methane molecules. The network with the global average pooling layer (left), and the network with the global max pooling layer (right) are shown. Both networks start by extracting data from QM7-X data set <cit.>, and converting it into graph representation. Each convolution layer (indicated by Conv$-l$) uses the trained weight matrix, $W_{l}$, and a relevant adjacency matrix $A$ to update the feature vector $H_{l}$ using Eq. (<ref>). The initial feature vector $H_{0}=X$ is shown with the magnitudes of the PBE0 atomic forces for the two methane configurations. Note that 3 convolution layers (of size $32$) are used in the left network, compare to the 4 convolution layers (of size $16$) in the right one. As the output scores indicate, both networks are successful in classifying the optimized and displaced configurations. | 2108.09637_2.jpg |
arxivcap_1111.0430_3 | Title: Fermi-LAT spectral analysis of Fermi, Planck, Swift and radio selected samples of AGN | Caption: SED of PKS 1502+106 from the $\gamma$-ray sample <cit.>. Red filled points: simultaneous multi-frequency data; green points: $\gamma$-ray data integrated over a period of 2 months centered on the Planck observing period, or ground-based data taken quasi-simultaneously; blue points: $Fermi$-LAT data integrated over 27 months; light gray points: literature or archival data. | 1111.0430_3.jpg |
arxivcap_2202.00203_8 | Title: Constraining low redshift [CII] Emission by Cross-Correlating FIRAS and BOSS Data | Caption: Cross-power fits for the line and continuum amplitude from , showing the CMASS analysis on the left and the LOWZ analysis on the right. The top row shows the MCMC contours of our fit to the data. Dark blue and light blue regions represent the 68 and 95 percent contours, respectively. The MCMC analysis uses a simple flat prior that restricts both $b_{\rm \cii}I_{\rm \cii}$ and $b_{\rm \cii}dI_{\rm CIB}/dz$ to positive values. This prior has a minimal effect on our best-fit $b_{\rm \cii}I_{\rm \cii}$ values, but it serves to prevent nonphysical negative values from counting towards our upper limit constraints. The bottom row shows red histograms of the distribution of best fit $\chi^2$ for 10,000 simulations in which we draw $a_{\ell m}$ amplitudes from Gaussian distributions for the cosmological signal, galaxy shot noise, and FIRAS auto-power (for more details of the simulation, refer to the parametric simulations described in Appendix <ref>). The $\chi^2$ of the maximum likelihood fits to the data are plotted as vertical dashed blue lines. | 2202.00203_8.jpg |
arxivcap_1912.05084_11 | Title: Bayesian Copula Density Deconvolution for Zero-Inflated Data in Nutritional Epidemiology | Caption: =10pt Graph depicting the dependency structure of the model developed in <cit.> for one episodically consumed component and one regularly consumed component. While other parameters are suppressed, the Box-Cox parameters $\lambda_{1}$ and $\lambda_{2}$ are shown to highlight nonlinear transformations of the surrogates. Compare with our model depicted in Figure <ref> in the main paper. | 1912.05084_11.jpg |
arxivcap_1802.01492_8 | Title: Analysing the Degree of Meshing in Medium Voltage Target Grids - An Automated Technical and Economical Impact Assessment | Caption: Cost comparison for example grid area: costs for all 50 separately optimized grids (left) and cost minimal reference grid (right) for each concept | 1802.01492_8.jpg |
arxivcap_2005.06146_6 | Title: Neutron spin resonance in a quasi-two-dimensional iron-based superconductor | Caption: {Superconducting properties of KCa$_2$Fe$_4$As$_4$F$_2$ crystals}. a, Temperature dependence of the resistivity within the $ab-$plane, the data is normalized by the resistivity at 300 K. Insert shows the superconducting transition at $T_c=33.5$ K. b, Magnetic susceptibility measurements by zero-field-cooling (ZFC) and field-cooling (FC) methods with $H // ab$. | 2005.06146_6.jpg |
arxivcap_1806.08658_5 | Title: Privacy-Preserving Identification via Layered Sparse Code Design:
Distributed Servers and Multiple Access Authorization | Caption: The relation between probability of correct identification and a) sparsity ratio, b) encoding rate. | 1806.08658_5.jpg |
arxivcap_2202.01503_8 | Title: Global sensitivity analysis based on Gaussian-process metamodelling for complex biomechanical problems | Caption: First-order and total-order Sobol indices and 95% confidence intervals (CI) for an increasing number of training samples. We use $M = \num{10000}$ Monte-Carlo samples, $\NGP = 500$ realisations of the Gaussian process, and $B = 300$ bootstrap samples. Subfigure A shows Sobol indices with metamodel CI and the sum of metamodel and Monte-Carlo CI for the six input parameters separately. Monte-Carlo abbreviates as MC. Subfigure B details the metamodel CI. | 2202.01503_8.jpg |
arxivcap_1906.05326_14 | Title: Magnetism and superconductivity in Fe$_{1+y}$Te$_{1-x}$Se$_x$ | Caption: Constant-energy scans of 6 meV through (0.5, 0.5) along [1$\bar{1}$0] direction at 100 K (blue circles) and 300 K (red diamonds) for (a) Cu0, (b) Cu02, and (c) Cu10 (see text). Lines through data are fits with Gaussian functions. In the inset of (c) we plot the ratio (R) of enhancement on the 100-K integrated intensities (I$_{100K}$) to that of 300 K (I$_{300K}$) for these scans at different energies, with $R = (I_{\rm 100\,K} - I_{\rm 300\,K})/I_{\rm 300\,K}$. The line through the triangles is a guide to the eye. Reprinted with permission from Ref. <cit.>, 2013 by the American Physical Society. | 1906.05326_14.jpg |
arxivcap_1109.6459_1 | Title: Folding Kinetics of a Polymer | Caption: Free energy $F_C(E) = E - T \ln(g(E))$ (black) and canonical probability function $P_C(E) \propto g(E) \exp(-E/k_BT)$ (green) obtained from WL simulation. Vertical lines schematize the interfaces $\lambda_i^B$ in folding (red) and the interfaces $\lambda_i^A$ in unfolding (blue) directions. Displayed also are energy distributions of states in the crystalline phase (blue) and in the globule phase (red); these graphs must be scaled according to the folding and unfolding rates <cit.> to be comparable with $P_C(E)$. The vertical axis of the probability function is not shown; these functions are normalized. The data are for the chain with $\chi = 1.07$, $T = 0.498$. | 1109.6459_1.jpg |
arxivcap_2210.09493_7 | Title: Strong lensing constraints on primordial black holes as a dark matter candidate | Caption: Product of posterior distributions based on image positions and priors of the three lenses B1422+231, PS J1606-2333, and WGD J2038-4008. As in Fig. <ref>, the selected "true" flux ratios used to obtain each distribution are $M_{\mathrm{PBH}} = 10^{4.1}$M$_\odot$, $f_{\mathrm{PBH}} = 0.02$ on the left and $M_{\mathrm{PBH}} = 10^{5.9}$M$_\odot$, $f_{\mathrm{PBH}} = 0.48$ on the right. | 2210.09493_7.jpg |
arxivcap_1104.3706_5 | Title: Detection of a large massive circumstellar disk around a high-mass young stellar object in the Carina Nebula | Caption: The 32 best fitting Robitaille SED models compared to the measured fluxes of the disk object. The best fit is shown as a solid line; the dashed line shows the stellar spectrum of the best fit model with the effect of the foreground extinction of $A_V = 5.59$ mag. The grey lines show the other good models as discussed in the text. | 1104.3706_5.jpg |
arxivcap_1902.04007_5 | Title: Scaling laws for harmonically trapped two-species mixtures at thermal equilibrium | Caption: Left: Plot of the total interaction energy versus time for a given equal number of particles in the two systems. In the inset the inverse temperatures of the two systems are shown versus time, confirming the presence of a regime in which thermalization is assured, and showing also the presence of an exothermic equilibration. The simulations were performed with $N_A=N_B=400$ particles, $\gamma$=20.0, $\lambda$=0.1, $\beta_A$=2.0, $\beta_B=0.2$, $m_A=m_B=1.0$, $\omega_A=1.0$, $\omega_B=144/89$. Right: Same quantity for varying number of particles in the two systems, qualitatively showing that the interaction energy increases with the number of particles, and that its fluctuations after the thermalization stage decrease with the number of particles. | 1902.04007_5.jpg |
arxivcap_2305.05311_2 | Title: Structured Sentiment Analysis as Transition-based Dependency Parsing | Caption: The proposed neural architecture and decoding steps for partially producing the dependency graph in Figure <ref>. Please note that stands for transition Attach-to-k. | 2305.05311_2.jpg |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.