dedup-isc-ft-v107-score
float64 0.32
1
| uid
stringlengths 32
32
| text
stringlengths 0
32.5k
| paper_id
stringlengths 1
14
| original_image_filename
stringlengths 5
222
|
|---|---|---|---|---|
0.44381 | 357bca555e2e4a5aa989abdf3ec6bbf0 | Zero temperature PMFRG flow of the static local and nonlocal susceptibilities χ_ij for the antiferromagnetic Heisenberg dimer. The grey dotted line represents the exact physical (Λ=0) result. | 2012.14836 | Dimer_chiflow.jpg |
0.511175 | 9d71b83a82bc4352be9c9104057ae35f | Z_2-invariant Majorana FRG flow equations for the interaction correction to the free energy (a), the self-energy (b), and the local (c) and nonlocal (d) four-point vertices. The order of labels 1=(α_1,_1) always corresponds to that on the left hand side of the vertex flow equations such that the site index is conserved along solid lines. In these equations, internal lines correspond to fully dressed Green's functions G_i(1,2), while the single scale propagator S_i(1,2) is represented by a slashed line. Similarly, the crossed line in a) corresponds to the local propagator [S G^0 G^-1]_i(1,2). | 2012.14836 | Floweqs.jpg |
0.480878 | da0fd373edb1488292c68acfa15d5d0a | (Free) energy, heat capacity per spin and static susceptibilities of the Heisenberg dimer with J=1 obtained via PMFRG (red symbols) at Λ=0 as a function of temperature. Displayed in solid (dashed) grey lines are the results obtained by (pseudo-fermion) exact diagonalization, as well as the finite temperature spin susceptibilities of the PFFRG in black symbols. Each data point corresponds to a fully converged flow with respect to Λ as demonstrated in the exemplary plot at T=0.1 (cf. <ref>). | 2012.14836 | TSweepDimer.jpg |
0.534986 | a602251bbd0244de95dda5d946b89963 | PMFRG results for the Heisenberg hexamer in analogy to <ref>. The corresponding PFFRG and PFED results of the spin susceptibility are included in the inset. | 2012.14836 | TSweepHexamer.jpg |
0.458286 | 02e7e28fc1424f199a976819b07a6cda | PMFRG results (dots) for the J_1-J_2 square lattice Heisenberg model at J_2/J_1=0.5. The panels depict the single site contribution to the interaction correction to the free energy, internal energy, heat capacity and uniform susceptibility (top to bottom). The HTSE data (dashed line, up to 9th order) is taken from Ref. <cit.>, its 4,5 Padé approximant is shown as a solid line. The iPEPS result for the ground state energy E_0/N=-0.495 from Ref. <cit.> is indicated as a dotted line. | 2012.14836 | Tsvelik-pm-fRG_square_J205.jpg |
0.490994 | 3ff11109d26c471bb534cf5df8f6a8cf | PMFRG results (dots) for the nearest-neighbor triangular lattice Heisenberg model. The observables presented are analogous to Fig. <ref>. The HTSE data (dashed line, up to 12th order) is taken from Ref. <cit.>, its 6,6 Padé approximant is shown as a solid line. | 2012.14836 | Tsvelik-pm-fRG_triang.jpg |
0.419837 | d2250e3ea46e428ea02a9ca782381bda | measurements available from public DES-SN3YR data is compared with our PS1 measurements. | 2012.14839 | crcheck.jpg |
0.439046 | 781d8c1780cc48b7bca97fe43ca8a545 | PS1 image stamps of three low-z SN Ia host galaxies with surface brightness magnitudes (m_SB) indicated on each panel. The circle is at the location of the SN Ia and indicates the aperture used to measure m_SB. | 2012.14839 | hostgalaxies.jpg |
0.436263 | af0d2d1a20be4a8693de1fc625de938a | Weighted HR vs. i-band m_SB with 1 mag bin size. The bins with m_SB = 16-17 and m_SB = 26-27 contain one event. | 2012.14839 | hrsb.jpg |
0.475157 | 7d908ee8b89a42d7a89aed773f1a529c | Correlation between our g-band SB and the local mass measurements from <cit.>. Vertical line indicates median m_SB for the low-z sample. Orange and yellow lines indicate median local mass for the bright and faint halves, respectively. | 2012.14839 | lcmass_sb.jpg |
0.466807 | b9fda383d5fc421b97ef82992b17f452 | 2012.14842 | Mother_feeding.jpg |
|
0.434366 | 5b7a1b3bb072430998a613c579e2f9e6 | 2012.14842 | after_birth_env.jpg |
|
0.416528 | 890f69e21a5443e8badde8e69b8c64d6 | 2012.14842 | agent_vision_central_in_focus.jpg |
|
0.395013 | 4ec77fcc08f146d9acd3bda8949c6309 | 2012.14842 | agent_vision_central_out_focus.jpg |
|
0.44619 | 5a92db40fb9b461585c522ad6b6d9d1d | 2012.14842 | agent_vision_far_peri.jpg |
|
0.441165 | 5920bfcad5f047fc8607b29f61c203e7 | 2012.14842 | agent_vision_near_peripheral.jpg |
|
0.435951 | 85613a6e829d4a17989a424fa6d9e280 | 2012.14842 | fetus_env.jpg |
|
0.476491 | 47f938d2bd8b465cac9d5b3b0c9f7dca | 2012.14842 | left_central.jpg |
|
0.460635 | 8ad9f41e337c40518125e77943030e87 | 2012.14842 | left_peripheral.jpg |
|
0.463245 | 538352274b4047d18f2659bcb407da03 | 2012.14842 | mother_showing_toy.jpg |
|
0.392863 | 001dd0ad35d64fb681d2afadb80cc40a | 2012.14842 | paper_rod.jpg |
|
0.442487 | b7a3f5aad2a34f9cb990f7f0a8b0486f | 2012.14842 | right_central.jpg |
|
0.437591 | 6730ddd87a224df085c62d027823def5 | 2012.14842 | right_peripheral.jpg |
|
0.42612 | 319823269d8746b5bb5dbd9bff6a6133 | 2012.14842 | rod_and_box.jpg |
|
0.452418 | f5306d2e16514b3f92cdb1810dc0e765 | Boxplots for empirical coverage of entrywise confidence interval CR_ijk | 2012.14844 | boxplot.jpg |
0.436026 | 5291079746a747d2af70990a1efd8eab | An example of Entity Discrimination (ED) task. For an entity pair with its distantly supervised relation in the text, the ED task requires the ground-truth tail entity to be closer to the head entity than other entities. | 2012.15022 | EP.jpg |
0.462709 | ecc3592b4dde4a409fe42a7d71a97eba | An example of Relation Discrimination (RD) task. For entity pairs belonging to the same relations, the RD task requires their relation representations to be closer. | 2012.15022 | RP.jpg |
0.424928 | f47bbfae69b1436b93a585f897b8151e | An example for a document “Culiacán”. All entities in the document are underlined. We show entities and their relations as a relational graph, and highlight the important entities and relations to find out “Where is Los Mochis”. | 2012.15022 | example.jpg |
0.427557 | 6cef1248812c473d94c6943e6cdbea8b | Impacts of relation distribution shifting. X axis denotes different ratios of relations, Y axis denotes test IgF1 on different partitions of DocRED. | 2012.15022 | rel.jpg |
0.409207 | c3da4faeb0004b89b8198bdefd9fb817 | Impacts of pre-training data's size. X axis denotes different ratios of pre-training data, Y axis denotes test IgF1 on different partitions of DocRED. | 2012.15022 | size.jpg |
0.461722 | 764bfc4e53294293acd080540cbfab78 | t-SNE plots of learned entity and relation embeddings on DocRED comparing BERT and _. | 2012.15022 | tsne.jpg |
0.425638 | 766e531a33004221820c462931cf95f1 | AGDN-HA layer Architecture. | 2012.15024 | 1.jpg |
0.466344 | 953b4b6068704c63b15e69ba6f17dbc8 | Single layer with multi-hop aggregations. With the same depth of receptive field in Figure <ref>, only one linear transformation is used. Moreover, multiple layers with multi-hop aggregations can be stacked in order to ensure higher model complexity. | 2012.15024 | AGDN_v3.jpg |
0.514718 | c480b97b3afb45b59b1fc2936e166a79 | Multiple MPNN layers. The dotted curves correspond to receptive field. The change of colors corresponds to a linear transformation. In MPNNs, the number of linear transformations is exactly the number of aggregations. | 2012.15024 | MPNN_v3.jpg |
0.531604 | aa70887546a0468ebe5633e5395c8e82 | The second protocol proposed for the closed 4-level system via local fields with opposite orientation. | 2012.15026 | 4states.jpg |
0.439449 | 105211e83dac414e9a1d80986909bb2a | Work W(t) (left) and power P(t) (right) after a quantum quench of the anisotropic XY model (<ref>) for N=1000 spins, with the anisotropic parameter η = 0.5. The battery internal (unquenched) Hamiltonian has zero field, h^(1) = 0. Colors label various quenching field h^(2) at both non-critical (black dashed and blue dotted) and critical values (red solid). | 2012.15026 | SpinChain.jpg |
0.459704 | c4450a08006a4bf2bb7b1968263dca2d | Bandwidth-Oriented Bounds | 2012.15027 | bandwidthBound.jpg |
0.446546 | 049406ef91bb44b2b770df5519e679bf | Storage-Oriented Bounds | 2012.15027 | storageBound.jpg |
0.461624 | 4db7379c12174fddab772efc51bbd64c | Time-Oriented Bounds | 2012.15027 | timeBound47.jpg |
0.380112 | 09528f26b3624c3eab59445c8d1128f6 | PSNR at different computational cost and parameter amount of our method and previous methods in SIDD <cit.>. The proposed NBNet achieves SOTA performance with a balanced computational requirement. | 2012.15028 | flop-psnr.jpg |
0.445807 | 4e6328c632f640b9ba4332a0caeb5baa | Denoising via subspace projection: Our NBNet learns to generate a set of basis for the signal subspace and by projecting the input into this space, signal can be enhanced after reconstruction for easy separation from noise. | 2012.15028 | projection_concept.jpg |
0.46117 | b812269e20ca430595c5284805383ffe | Temporal average of the number of detected vortical structures over all planes using (a) λ_ci criterion with Gaussian filters for two kernel sizes of 3×3 and 5×5, and (b) Γ_2 for the shear-dominant votices (|Γ_2|≤ 2/π), and rotation-dominant vortices (|Γ_2|≥ 2/π). | 2012.15031 | Gamma2_counts.jpg |
0.515043 | 97ce114af7d4425292854c0bc9a6a5dd | Instantaneous vortex core structures identified by Γ_2 scheme with a 3×3 kernel size corresponding to 0.12D_B. (a) Plane PL-1 at Re_p=100, (b) Plane PL-3 at Re_p=272, (c) Plane PL-4 at Re_p=270, (d) Plane PL-6 at Re_p=528, (e) Plane PL-6 at Re_p=736, and (f) Plane PL-6 at Re_p=948. Computed values are for the whole porous bed. | 2012.15031 | Gamma_2.jpg |
0.42991 | 5789f752f0c54394b77851d2226ee9cd | The variation of time averaged pore-scale vortex strength as a function pore-scale Reynolds number (Re_⟨ p⟩) non-dimensionalized by pore-averaged longitudinal time scale ⟨τ_L_⟨ p⟩⟩. (a) Time averaged circulation _⟨ p⟩⟨τ_L_⟨ p⟩⟩, and (b) time and pore-averaged vortex strength ⟨λ_ci_⟨ p⟩⟩⟨τ_L_⟨ p⟩⟩. | 2012.15031 | Gamma_rms_local_3.jpg |
0.457478 | bab9653ffb104080a4d3a47cd13d31f4 | Vortex size in the entire porous bed: (a) the variations of vortex size computed based on longitudinal integral length scale (⟨ L_L⟩) normalized by D_H, (b) vortex size computed based on vortex boundary scheme (Γ_2) for the entire vortex core size (|Γ_2|≤ 1), shear-dominant votices (|Γ_2|≤ 2/π), and rotation-dominant vortices (|Γ_2|≥ 2/π), and (c) vortex size based on λ_ci (L_λ_ci), as well as λ_cr/λ_ci criteria. Computed values are scaled by hydraulic diameter D_H for the whole porous bed for the cases of Reynolds numbers given in [tab:cases]Table <ref>tab:cases . | 2012.15031 | L_Lambdaci_Lint_G5.jpg |
0.465531 | 3b431dfa6e96477eb4703561a0e97c6a | Contours of the instantaneous swirl strength of detected vortical structures normalized by the longitudinal Eulerian integral time scale (λ_ciτ_L) overlayed on the LES filtered velocity with Gaussian kernel of 3×3 plotted for a time step of T = 20Δτ for all six cases of [fig:time-avg]Figure <ref>fig:time-avg. (a) Plane PL-1 at Re_p=100, (b) Plane PL-3 at Re_p=272, (c) Plane PL-4 at Re_p=270, (d) Plane PL-6 at Re_p=528, (e) Plane PL-6 at Re_p=736, and (f) Plane PL-6 at Re_p=948. | 2012.15031 | L_ci_bed_new_cont.jpg |
0.405339 | f8b2e28f303d417c940ff2428a95fedc | Variations of vortex size as a function microscopic Reynolds number (Re_⟨ p⟩) normalized by D_H_⟨ p⟩. (a) The variations of vortex size computed based on pore-averaged longitudinal integral length scale normalized by D_H, (b) vortex size computed based on vortex boundary scheme (Γ_2), and (c) vortex size based on λ_ci (L_λ_ci) criteria. | 2012.15031 | L_local_size.jpg |
0.431151 | 8933d6a3aeba42a895b0cae8308bc0d1 | Temporal average of number of detected vortical structures in the entire bed based on λ_ci criterion representing the number density of pore-scale vortical structures normalized by the area of pores (A_⟨ p⟩). | 2012.15031 | N_vortex_local_2.jpg |
0.435596 | d647a44fd2c44f7a86fe0165596d1c7e | (a) Porous medium with solid and fluid phase refractive indices matched, (b) sequence of images of laser light sheet distortion when traveling through an index mismatched porous bed for light-sheet wavelength of λ = 527 nm by applying these index mismatch values. | 2012.15031 | RIM_Bed_gray.jpg |
0.43955 | 4abebf611d58497a9348e54674ee9b71 | Schematic of the experimental flow loop and test section for PIV imaging together with the bubble control, and the bypass loop | 2012.15031 | Visio_low_Loop_System_Updated4.jpg |
0.427953 | e36463ae20ad4e319c37f15f69ba5ccf | The three-dimensional geometry of the randomly packed porous bed. The location of one of the image planes for PIV data collection is shown for reference, image axis is along coordinate Z. | 2012.15031 | bed5.jpg |
0.442813 | a548981499a64bdb846e9b5a63893fc6 | Schematic of the geometrical demonstration of individual pore-scale (local) regions for estimating the local Reynolds number Re_⟨ p⟩ versus global (macroscopic) for longitudinal (y-direction) velocity intensities normalized by interstitial velocity globally, and by pore-averaged velocity locally. Color bars represent the global and local velocity intensities. | 2012.15031 | foo_hot_2_high_res2.jpg |
0.415373 | 6e3aced0196a4c7387215dec422acb17 | (a)-(d) Contours of the instantaneous swirl strength of detected vortical structures normalized by spatially averaged longitudinal Eulerian integral time scale (⟨τ_L ⟩) overlayed on the LES filtered velocity with Gaussian kernel of 3×3 plotted for Re_p = 270 at four consecutive time intervals with the time-steps of 16.2 ms. | 2012.15031 | lambda_ci_time.jpg |
0.485458 | bdf32f04c7dc42ccbd003f7a9b7466c6 | Instantaneous λ_cr/λ_ci criterion indicating vortical structures with spiral motions passing through the plane of visualization at (a) Plane PL-1 at Re_p=100, (b) Plane PL-3 at Re_p=272, (c) Plane PL-4 at Re_p=270, (d) Plane PL-6 at Re_p=528, (e) Plane PL-6 at Re_p=736, and (f) Plane PL-6 at Re_p=948. | 2012.15031 | lambda_cr_3.jpg |
0.417269 | 12be7f38744645c29a5b0ce5b496d412 | The variation of vortex strength versus Re_p normalized by (D_H/V_int) (a) normalized vorticity rms ⟨ω_rms⟩ D_H/V_int, and (b) normalized time- and space-averaged vortex strength ⟨λ_ci⟩ D_H/V_int. | 2012.15031 | omega_rms3.jpg |
0.436132 | 66eefdcc296a4a799df40b9b2af95f7c | The vector-map configuration of instantaneous velocity at different planes in the experimental apparatus for Reynolds numbers ranging from 100 to 948. Six different test cases were obtained. | 2012.15031 | planes8.jpg |
0.475809 | 2957f9d3f4534186820ad3093ab650e5 | Contours of non-dimensional longitudinal velocity V/V_int overlayed with time-averaged normalized vector-map configuration of four different planes for Reynolds numbers ranging from 100 to 948; (a) Plane PL-1 at Re_p=100, (b) Plane PL-3 at Re_p=272, (c) Plane PL-4 at Re_p=270, (d) Plane PL-6 at Re_p=528, (e) Plane PL-6 at Re_p=736, and (f) Plane PL-6 at Re_p=948. | 2012.15031 | time_avg44.jpg |
0.403232 | 2abc94853aa949df8df94274a49638a9 | Pore-scale contours of non-dimensional longitudinal velocity V/⟨V_⟨ P⟩⟩ overlayed with the time-averaged normalized vector-map of four planes for Reynolds numbers ranging from 100 to 948; (a) pore B in plane PL-1 at Re_p=100, (b) pore B in plane PL-3 at Re_p=272, (c) pore B in plane PL-4 at Re_p=270, (d) pore A in plane PL-6 at Re_p=528, (e) pore A in plane PL-6 at Re_p=736, and (f) pore A in plane PL-6 at Re_p=948. | 2012.15031 | time_avg_pore.jpg |
0.403532 | 4429ee7c8a2a4f0fb50c2bb593e99208 | (a)-(d) Sequences of instantaneous out-of-plane vorticity contours of detected vortical structures normalized by spatially averaged inverse longitudinal Eulerian integral time scale (⟨τ_L ⟩) plotted for Re_p = 270 (PL-4) at four consecutive time steps with interval of 16.2 ms. | 2012.15031 | vor_porB_time_2.jpg |
0.453863 | 36ef49163f7b478888a4e5fac0a7b41d | Continuum parameters and fluxes are plotted for all spectra. In descending order, we plot values for the two column densities, the reflection fraction, the 1-10 keV absorbed and unabsorbed fluxes of the central engine (F_cent and F_cent, unabs), the unabsorbed to absorbed flux ratio (F_cent, unabs/F_cent), and the Eddington fraction calculated using the 1-10 keV unabsorbed flux. The points in orange correspond to spectra where a flare was observed in the light curve. The red line in the first panel denotes the measure of one Thompson optical depth. | 2012.15033 | fig3.jpg |
0.525591 | c09b8d263d904846b91eeecd6e79540c | Correlations between internal obscuration and F_ unabsorbed, central, reflection fraction, and scattered gas are shown. All values are in logspace. The Spearman Rank-Order correlation coefficient, r_s is shown for each correlation, along with the probability of false correlation. | 2012.15033 | fig4.jpg |
0.433074 | ba81691d750346e988206e46dc5aa05f | Histogram of column densities and reflection fraction. The y-axis refers to the number of observations. The first few observations (which roughly correspond to the first month of the monitoring campaign) have lower obscuration values and smaller reflection fractions, similar to the flares. | 2012.15033 | fig7.jpg |
0.462861 | 4e5afcb576584921a6e8b0442bfa7914 | Contour plot showing column density vs. reflection fraction for observations 00034292227 and 00034292228, taken one day apart. The column densities are distinctly different up to 2 σ error bars, justifying the high values for internal obscuration returned by some of the fits. | 2012.15033 | fig8.jpg |
0.445269 | b7057f39171349bd91c2164698ee6073 | Human Learning Before AI | 2012.15035 | baduk_learning.jpg |
0.495763 | 5252961757c4431e8c5118fe1d9ba694 | Human Learning After AI | 2012.15035 | baduk_learning_after.jpg |
0.431065 | 196dc650b637491eac272f64a8f14cd1 | Histogram of Human-AI Gap from a cheating player | 2012.15035 | eunji_dist_of_human_ai_gap_means.jpg |
0.501844 | 35ebaa681a624a3faeb1431b23f6557d | Mean Human-AI Gap over the Course of Match | 2012.15035 | human_ai_gap_across_moves_exact_dates_v3.jpg |
0.523688 | d31be5a1c4d64e1eb3227dcaf96bba48 | Change in Gap between AI and Humans (Δ) over Time (Control group) | 2012.15035 | military_coefplot.jpg |
0.451455 | b51beab008f64f2597a2c604feed4427 | Human-AI Gap over Time | 2012.15035 | treatment_coefplot.jpg |
0.461249 | 5c2f1575ff5b4c5aa85a7ac292d3ca11 | Evolution of four parameters from ReLU activation for SGD and PDE algorithms varying Δ. The x-axis represents the iteration, where 1-9 represents iterations 1-9, 10-19 represents iterations (10,20,30,40,50,60,70,80,90,100), 20-28 represents iterations (200, 300, 400, 500, 600, 700,800,900,1000), and so on. | 2012.15036 | IterationsPlot.jpg |
0.458213 | 89ccd510cf8040b3b43e46314de6bdc7 | R(2ρ;1ρ) visualized in neural network. | 2012.15036 | JointDistributionVisualization.jpg |
0.452738 | 11c9c50ffdf44b07b239f5f9d6c5c31d | 1R(1ρ) visualized in neural network. | 2012.15036 | MarginalDistribution1Visualization.jpg |
0.473277 | 7dc07ec2aeae4de6ad66064735c4cd4a | 2R(2ρ) visualized in neural network. | 2012.15036 | MarginalDistribution2Visualization.jpg |
0.447016 | c57de66168b24d27b076f19b586be5ad | Proportion of Missing Data in Training Set. | 2012.15036 | MissingData.jpg |
0.432733 | 0f544722bb6c4699aa696c495424ab18 | Evolution of weight distribution under PDE approximation with r(0) taken from extracted weights of single hidden layer neural network with 100 nodes. | 2012.15036 | PDEEvolutionSingleLayerNetwork.jpg |
0.358867 | 0cd8b30ab9b4488ba0b114be86eee239 | Table summarizing internal data set features from American Family. | 2012.15036 | Table1.jpg |
0.413083 | 42c7a2d1f4a9454ba2547d9b8aa5c199 | Weight Distributions after running PDE approximation on extracted weights from neural networks with one hidden layer of 100 nodes and two hidden layers of 100 nodes each, respectively. | 2012.15036 | TwoTheoreticalNetworksAfterPDE.jpg |
0.413016 | 114bccd14ffe41eea843d82713245f2e | Variance explained from MCA. | 2012.15036 | Var_Explained.jpg |
0.496844 | 1d3362801a494008bdc3d0e86b8b03d5 | Second layer weight distribution observed on two-layer network (250,250) ran on AmFam Data, and weight distributions derived from PDE simulation using Piecewise Linear Activation Function for various choices of weights from second layer. | 2012.15036 | WeightDistribution_2Layer.jpg |
0.41522 | 23cf097094cb4106b175e9261734e005 | Discriminators in multi-adversarial learning. | 2012.15037 | adver_train.jpg |
0.416272 | ca08d7f7563d4a6d9633e4232dd524bd | Visualization of the learned attention weight of neighboring air quality and weather stations. | 2012.15037 | casestudy.jpg |
0.424507 | 133610e5d60e4412870e6ab5aa2449e7 | Spatial distribution of air quality and weather monitoring stations in Beijing. Two types of stations are monitoring exclusively different but correlated air quality and weather conditions in different city locations. | 2012.15037 | distribution.jpg |
0.462094 | d3a186a9759b42e18a2907c4447cce7d | An overview of MasterGNN. | 2012.15037 | framework.jpg |
0.470636 | 4133d59d0b7a42b3beeee73e14ce6823 | LSTM (cell) | 2012.15041 | LSTMGrey.jpg |
0.471063 | 9654eae0db794b9e896683375b06c7a0 | RNN process | 2012.15041 | RNNGrey.jpg |
0.503602 | 3b63391262fd42d9ad2a5473ecc34975 | Accuracy variation during learning time (100 Epochs) | 2012.15041 | accuracy.jpg |
0.49862 | 805e07e9681749198e91276c5c1eefd6 | Area under the curve (AUC) variation during learning time (100 Epochs) | 2012.15041 | auc.jpg |
0.428187 | 07bc2384170d4b668efcee49a1527351 | Loss variation during learning time (100 Epochs) | 2012.15041 | loss.jpg |
0.428866 | 755070465f464f4ca5205cac500b5ca4 | Precision variation during learning time (100 Epochs) | 2012.15041 | precision.jpg |
0.496009 | d27f523f73134c60abb7fad3aa90257e | Recall variation during learning time (100 Epochs) | 2012.15041 | recall.jpg |
0.414302 | bf41d6adaec1487e9ede4999e2514ede | Snapshots, recorded during the evolution of the model system, having active particles in hydrodynamic environment, are presented from different times. Locations of only the active particles are marked. These and other results for nonzero activity correspond to f_a=3. | 2012.15043 | fig1.jpg |
0.463914 | df496ee5be7a4f189ca75726b7465be9 | Plots of the average mass, M, of the clusters versus time (t), are shown on a log-log scale. We have included data that were obtained in presence of hydrodynamics (a) and in absence of hydrodynamics (b). Results for both passive (f_a=0) and active (f_a=3) cases are shown. The dashed lines represent power-laws with mentioned values of the exponents. | 2012.15043 | fig2.jpg |
0.442119 | 65f06b0d11a5446f88b38f62976e2303 | (a) Plots of masses in two clusters, one each from f_a=0 and f_a=3, are shown as a function of time. These were obtained by following the cluster trajectories for extended periods of time. (b) Log-log plots of -dn/dt versus n, for f_a=0 and 3. The dashed lines represent power-laws, exponents for which are mentioned next to them. These results correspond to the situation when hydrodynamics was on. | 2012.15043 | fig3.jpg |
0.504549 | 34742dff328a44659c921c2d5e3f1ed1 | (a) Plots of mean-squared-displacement (of clusters) versus time, on a log-log scale. Typical results for passive and active cases have been included. The dashed lines represent power-laws. Corresponding exponents have been mentioned in appropriate places. (b) Root-mean-squared velocity of clusters is plotted versus average mass of clusters, on a double-log scale. For comparative purpose data from another model (AM), showing z=-1/2, expected for uncorrelated motion of particles, are also included. See Ref. <cit.> for the details on this model. These results correspond to hydrodynamic environment. | 2012.15043 | fig4.jpg |
0.498203 | c4fc76064cfb490e98688be6287272ec | The total thermal conductivity, κ (square symbols connected by a solid guide to the eye), the electronic component κ_el computed from ρ assuming the validity of the WF law (solid lines), and the “non-electronic” component, Δκ, (dashed lines) as a function of T. The two lattice directions are indicated by red for c axis and blue for a. The grey bar indicates a critical region where | 2012.15048 | All_kappa_5.jpg |
0.475679 | 99a058f2c5324d82ad0ee15e23255579 | A typical phase delay as a function of heating laser modulation frequency, ϕ(ω_0)(red dot and error bars) obtained by sweeping ω_0. This particular set is measured on ErTe_3 along the a axis at a separation distance of 16.8 μm at 300K. The blue line is a fit using the form in Eqn. <ref>. | 2012.15048 | ErTe3_PhaseFitAaxis300K16-8um_4.jpg |
0.444782 | 70d1b9bc00c141328f42c2e60efd85b1 | (color online) The schematic shows the optical paths of the setup. (a) Path of the heating laser (b) Path of the probing laser. The reflected light traverses the same path before gathered by a photodetector. (c) Path of illumination light source and camera vision. (d) Focused laser spots at typical measurement separation distance. The screenshot is taken on a just cleaved Pd_0.003ErTe_3 sample surface for visual interest, where atomically flat terraces in the bottom of the image can be seen. The surface of crystals measured in the main text are mostly featureless. | 2012.15048 | Microscope.jpg |
0.401237 | e69afdfb36d14a5d8b1b49759cd61455 | normalized density profiles for isothermal SIDM halos near a black hole for two different values of the 1D velocity dispersion. For the region outside the annihilation radius r ≫ r_ann, the density profiles follow a power law of r^-3/2. Inside the annihilation cusp, ρ(r)∝ r^-1/2. If dark matter does not annihilate, the density could increase exponentially towards the black hole (dashed). | 2012.15050 | density.jpg |
0.46782 | 74d23481e39b4fdcb9bd486ea8de1141 | The expectation of the DM distribution, ⟨ DM ⟩, as a function of survey fluence limit, F_0, for an homogeneous IGM. We use the best-fit energy power-law index, γ̂, for the constrained spectral index of α=1.5 and for the three redshift evolutionary models listed in Table <ref> via eq.(<ref>). The blue curve represents no redshift evolution (n=0), orange linear redshift evolution (n=1) and green quadratic redshift evolution (n=2) with respect to the CSFR. We extend the fluence survey limit down to anticipated SKA levels for reference purposes. Here, we use the Parkes response curve, as the corresponding ASKAP response-based curves are ostensibly the same, hence are omitted. | 2012.15051 | FluenceDMCurve.jpg |
0.450145 | 6a900ccbbbe04a4389168ae41e90e68b | DM response curves for the Parkes (blue) and ASKAP (orange) telescopes using eq.(<ref>) with the parameters of Table <ref> and for the population pulse-width mean μ=3.44 ms and standard deviation σ=2.66 ms. | 2012.15051 | ResponseCurves.jpg |
Subsets and Splits