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
|
|---|---|---|---|---|
0.452815 | 093db3de68214a55b02960ae69c11dd7 | Evolution of model parameters from 8th March–30th November. Median prediction shown as black line, alongside 95% CI (light grey) and 50% CI (dark grey). Horizontal dashed lines correspond to median, and 95% CI values for the parameter prior estimates. (a) transmission rate, β; (b) infectious period, τ; (c) case to hospital transition probability, κ_H; (d) hospital to critical care transition probability, κ_C; (e) hospital death probability, μ_H; (f) critical care death probability, μ_CC. | 2012.12390 | Figure_5.jpg |
0.46823 | 84b0b511f7b9441bb2595125b15f3a9c | Estimated prior distributions (black) and predicted posterior distributions (red in (a)–(d); multi-coloured in (e) and (f), to reflects the different weekdays, d(t)) for the model parameters: (a) latent period, α; (b) duration of case to hospital, λ_C; (c) duration of hospital to recovery, λ_H; (d) duration of critical case to recovery, λ_CC; (e) weekday correction, Δ_d(t); (f) weekday death correction, Δ^D_d(t). | 2012.12390 | Figure_6.jpg |
0.506033 | 77d1f83d0a6946d98202d4be7cb8df94 | Estimated prior distributions (black) and predicted posterior distributions (red) for overdispersion terms (ω_I, ω_H, ω_CC, ω_D) used in the negative binomial distributions. | 2012.12390 | Figure_7.jpg |
0.398171 | b363a15eb4ef4019a569aae00dbdb93d | First wave forecasts: Five 2-week forecasts for the second wave, from 8th March–17th May. Vertical blue lines indicate the date each forecast was produced. Median forecast values are shown by a black line, alongside 95% CI (light grey) and 50% CI (dark grey). Real-world surveillance data is shown in red, and all y-axes use a log-scale. (a) Reported cases; (b) hospitalisation; (c) critical care, and (d) daily deaths. Vertical blue lines show the position of the first point in each 2-week predictions and arrows point the direction of prediction. | 2012.12390 | Figure_8.jpg |
0.436638 | d940eca2c7274f868a0c4a676433c63e | Second wave forecasts: Five 2-week forecasts for the second wave, from 1st September–10th November. Vertical blue lines indicate the date each forecast was produced. Median forecast values are shown by a black line, alongside 95% CI (light grey) and 50% CI (dark grey). Real-world surveillance data is shown in red, and all y-axes use a log-scale. (a) Reported cases; (b) hospitalisation; (c) critical care, and (d) daily deaths. Vertical blue lines show the position of the first point in each 2-week predictions and arrows point the direction of prediction. | 2012.12390 | Figure_9.jpg |
0.46943 | 9d096f02280c487cad317339ea4fcc72 | Sketch of the phase diagram of model (<ref>) for density n=0.25. Four phases appear: a regular LL fermionic phase F, paired LL phases P_0 and P_π and a coexistence phase C with central charge c=2 where fermions and P_π pairs are mixed. | 2012.12393 | Fig1.jpg |
0.54451 | c007fd90cc1c46738b69563355773d03 | Absolute value of the Fourier transform of pair correlations for an open chain with L=80 and t^'/t=± 4. Open symbols are the t^'/t=4 data shifted by π. Inset: Pair correlations. | 2012.12393 | Fig2.jpg |
0.457547 | e672690acd42491395876dbf61b2e1fb | (a) First and (a') second derivatives of the energy per site e_0 as a function of t^'/t for three system sizes L=56, 136, 200. Dotted lines are predictions of the 2F model. Arrows point toward typical band structures of the 2F model for t^'/t=0.25, t^'/t=1.7 and t^'/t=4. (b) Fitted central charges as a function of t^'/t. (c) Single-particle kinetic energy K_f and pair kinetic energy K_b probing almost directly n_f and n_b. | 2012.12393 | Fig3.jpg |
0.397355 | c61c79fce4bf402a9b22d12404b8316b | Map of the absolute value of the Fourier transform of local density fluctuations ⟨δ n_j⟩ as a function of t^'/t for an open chain with L=200. | 2012.12393 | Fig4.jpg |
0.439178 | 9dc3f1bac0f64aa1ab9cfa09caa03ce4 | Optimal fermionic and bosonic densities obtained by minimizing e_2F/t for a total density n=0.25. | 2012.12393 | supmat_fig1.jpg |
0.434093 | 7421babcd9a44047b6deb0febaa2f6f4 | First and second derivative of ϵ_GS(τ,n_f(τ)) as a function of τ for a total density n=0.25. | 2012.12393 | supmat_fig2.jpg |
0.441463 | ec102c15f50c49369975fd34b36b1944 | Ground state energy of a system governed by H_t^' with PBC on a lattice of size L=40 as a function of the filling n∈ (0,1]. | 2012.12393 | supmat_fig3.jpg |
0.500846 | e5620f7bde0b4bfb91bc7d1372deeb9d | Absolute value of the fermionic occupation factor (blue dots) compared to its analytical estimate (orange line). The numerical parameters are L=80, t=1, t^'=4 with OBC. | 2012.12393 | supmat_fig4.jpg |
0.475471 | d44c2cc7e4324968953d38b827e610d5 | Fits of the entropy : symbols are DMRG data, black lines are fits with Eq. (<ref>). | 2012.12393 | supmat_fig5.jpg |
0.506898 | 6e32b08ae73d4862b2a9b24c54b659ea | Application | 2012.12394 | application.jpg |
0.472901 | d0df5b45a1da46688dcf5ed9b503d279 | Distance of outliers from P_Z mean increases. | 2012.12394 | boxplotPz_clean.jpg |
0.399901 | 1bfec14f3c214d0da709e5e22a9e14d8 | Average Precision on testing set for WALDO, WAE <cit.> and CoRa <cit.>. | 2012.12394 | convergence2.jpg |
0.396802 | 5e25884a99874cea96e2a6b15a1b5691 | Effect of positive training the discriminator D_γ on the residuals. Residuals are computed as ‖ G^I_θ(ẑ_i) - x_i‖^2_2 - ‖ G^O_ω(ẑ_i) - x_i‖^2_2. | 2012.12394 | figure_advantage.jpg |
0.396503 | 3626a73b52a1438cb26576ab71240979 | Generated Inliers and outliers from Cellphone and Laptop categories by randomly sampling data points from prior distribution z | 2012.12394 | generated_realimg.jpg |
0.483182 | a06585df29e94127bf77ff698a156ee7 | KDD99 Outlier Generation | 2012.12394 | kddoutput.jpg |
0.473536 | 273a485377154df294695fbc4e230a81 | In Figure (a), data lives on a linear manifold which can be inferred using PCA. The point A has high reconstruction error while point B has zero reconstruction error. Thus A will be reported as an outlier but B which is an outlier will be missed. In Figure (b), au autoencoder can learn the curved manifold and ideally B will have zero reconstruction error. The promise of Wasserstein distance is that because it is induced by a metric, both A and B will be reported as outliers. | 2012.12394 | motivate.jpg |
0.420528 | 1d9d2a2aca43437ab66525e67a1c5ada | Outlier detection of retail patterns. Outliers constitute a very small percentage of the data. | 2012.12394 | outliers_walmart2.jpg |
0.398631 | e39b44f2d6f44ac5b890e596475fa1e0 | Reconstruction of Inliers and outliers from Cellphone and Laptop categories. In each Figure, sampled images in the first row are original images; images in the second row are reconstructed images. | 2012.12394 | real_data.jpg |
0.444558 | 1fdcdaba45f74512b1e41a5e1d68e442 | Architecture | 2012.12394 | waldo_diagram.jpg |
0.420274 | 65e814f40ecd4b2986dffce8d9bd59b0 | Early fusion | 2012.12395 | l_v1.jpg |
0.390011 | 0401a05e7b204de9a8e11549f2554fe2 | Later fusion | 2012.12395 | l_v2.jpg |
0.39188 | 268895f819164932a62711dffc9f272c | (Color online) Calculated ground state energies in MeV using chiral NLO, and N^2LO interactions at Λ = 450 MeV (blue and green symbols) in comparison with experimental values (red levels). For each nucleus the NLO, and N^2LO results are the left and right symbols and bars, respectively. The open blue symbols correspond to incomplete calculations at N^2LO using NN-only interactions. Blue error bars indicate the NCCI extrapolation uncertainty. All results shown are for α = 0.08 fm^4. The light (coral) and dark (grey) shaded bars indicate the 95% and 68% DoB truncation errors, respectively, estimated using the Bayesian model C̅_0.5-10^650. | 2012.12396 | GSEnergies.jpg |
0.446436 | 1cc93dcf924b4dd29738392db90eec7c | (Color online) Results for the differential cross section, nucleon and deuteron analyzing powers A_y^n and A_y^d as well as deuteron tensor analyzing powers A_yy, A_xz and A_xx in elastic nucleon-deuteron scattering at laboratory energy of E_N = 70 MeV at NLO (yellow bands) and N^2LO (green bands) for Λ = 450 MeV. Dotted lines show the N^2LO results based on the SMS NN forces from Ref. <cit.> accompanied with the unsubtracted (i.e. with C(M_π) = 0) 3NF from Ref. <cit.>, while dashed-dotted lines are N^2LO predictions based on the SCS (NN+3NF) interactions from Ref. <cit.>. The light (dark) shaded bands indicate 95% (68%) DoB intervals using the Bayesian model C̅_0.5-10^650. Open circles are proton-deuteron data from Ref. <cit.>. | 2012.12396 | NdElastic_70MeV.jpg |
0.455828 | a483b6494f74460eab4ad9b44d8d3592 | (Color online) Results for the differential cross section, (a)-(d), and nucleon vector analyzing power A_y as functions of the kinematical locus variable S for the deuteron breakup reaction at laboratory energy of E_N = 65 MeV at NLO (yellow bands) and N^2LO (green bands) for Λ = 450 MeV. Proton-deuteron data for (a), (c), (f) are taken from Ref. <cit.> while those shown in (b), (d) and (e) are from Ref. <cit.>. For remaining notation see Fig. <ref>. | 2012.12396 | breakup.jpg |
0.434815 | 2a5bd265fc6749079f199a6c7b3d14ae | (Color online) Predictions for various Nd scattering observables based on the c_D and c_E values given in Eq. (<ref>). For every considered observable X, a solid dot corresponds to the ratio of its calculated value X_ th at N^2LO to the corresponding experimental value X_ exp. The smaller (blue) error bars correspond to the experimental relative uncertainty δ X_ exp/X_ exp, where δ X_ exp includes both the statistical and systematic errors added in quadrature. The larger (orange) error bars also take into account the estimated truncation error δ X_ th and correspond to √(δ X_ exp^2 + δ X_ th^2)/X_ exp. Experimental data are from <cit.> (2003 Schoen et al.), <cit.> (2001 Abfalterer et al.), <cit.> (2002 Sekiguchi et al.) and <cit.> (2005 Ermisch et al.). | 2012.12396 | cD_consistency.jpg |
0.412605 | 27510d184a7a4a71a5c7bd40bd553cec | (Color online) Correlation between the LECs c_D and c_E induced by the requirement that the ^3H binding energy be reproduced for the cutoff choices of Λ = 450 MeV (blue dashed line) and Λ =500 MeV (red solid line). | 2012.12396 | cDcE_correlation.jpg |
0.430225 | 6065f43bcda040a2ad6028df5827370a | (Color online) Central values (dots) for the excitation energies from Tables <ref> and <ref> with 95% Bayesian confidence intervals for truncation errors only indicated as error bars. (We omit the 0^+ in ^8Li because an experimental value is not available and the lowest 2^+ in ^12B because the experimental ground state is not correctly predicted.) For each excitation from the calculated ground state, the upper (blue) bar is for Λ = 450MeV and the lower (red) bar is for Λ=500MeV. All results shown are for α=0.08 ^4. | 2012.12396 | excitation_energy_ratios_both.jpg |
0.512717 | d72bec8de617494fa12cca8fe0578340 | (Color online) Prior pdf for the variance ^2 of the expansion coefficient with several choices of hyperparameters ν_0 and τ_0. | 2012.12396 | prior_figure.jpg |
0.471654 | 1a2a461423764e8c9382a2544cac24e8 | (Color online) Order-by-order 3 α breakup threshold of ^12C. Error bars indicate the NCCI extrapolation uncertainties only. | 2012.12396 | res_A12Z6N6_E3alpha_threshold.jpg |
0.451813 | 207a653bd6d24be090341ef539006972 | (Color online) Comparison of ground state energies of p-shell nuclei between chiral EFT calculations at N^2LO and experiment. All results shown are for α = 0.08 fm^4 and error bars indicate the NCCI extrapolation uncertainties only. | 2012.12396 | res_Egs_comparison_LENPIC-SCS-SMS.jpg |
0.496947 | 2792078fd4824bd7bce7004ee83fa04c | (Color online) Order-by-order excitation spectra of ^10Be (top-left), ^10B (top-right), ^12B (bottom-left), and ^12C (bottom-right). All excitation energies are obtained with SRG parameter α=0.08 fm^4; open symbols are with Λ=450 MeV, closed symbols are with Λ=500 MeV, and horizontal lines indicate experimental values <cit.> ; dashed lines indicate 0^+ states not in the low-lying spectra of our NCCI calculations. | 2012.12396 | res_spectra_A10A12.jpg |
0.457049 | c727ac907abe4f56b8c5abe11398872e | (Color online) Order-by-order excitation spectra of ^7Li (top) and ^8Li (bottom). All excitation energies are obtained with SRG parameter α=0.08 fm^4; open symbols are with Λ=450 MeV, closed symbols are with Λ=500 MeV, and horizontal lines indicate experimental values <cit.> . | 2012.12396 | res_spectra_Li7Li8.jpg |
0.455661 | 59c9983766bf4f93b6d8193470cd9383 | (Color online) Expansion coefficients for the individual energy levels in Tables <ref> and <ref> with α = 0.08 fm^4 and Λ=450 MeV. These are extracted according to Eq. (<ref>) with a fixed value of Q ≈ 0.31 and X_ ref taken from experiment <cit.> (or the N^2LO result for the 0^+ in ^8Li). | 2012.12396 | spectra_coefficents.jpg |
0.407797 | 9a63e51679084cb2ace42e1aa45d6b9d | For centralized training, data from each client is stored in the train and test set. For federated training, each vehicle is treated as a client and each local LSTM model is trained using local data. The client models are federated at the server to form a global model. Clients are randomly chosen for participation during the training step every round. (red x represents the vehicles not participating in that individual round, but eventually all will participate via multiple rounds) | 2012.12401 | doe.jpg |
0.510754 | ff20410563394e8497f05502d07c37ee | 2012.12401 | leftlanechangeresults.jpg |
|
0.45462 | f0b55d448386498c82ea6d6fe77aabeb | LSTM model architecture for turn signal prediction task. The output layer of the fully connected network has three states, each corresponding to left, right or off state of the turn signal. | 2012.12401 | lstm.jpg |
0.55693 | 62cf617898404b72b77d92b7856e27c0 | 2012.12401 | rightturnresults.jpg |
|
0.427819 | 7e96dfed7adf475b87746ee986e70719 | Trade-off between accuracy and model size by varying feature channel number C and block number N of the network. | 2012.12402 | param_num.jpg |
0.42245 | a9526fe7caa0412c8977a9dd8566c045 | ADTMPC produces consistent reductions in uncertainty over a variety of cyclic trajectories and BLSC controllers (left). ADTMPC is generalizable to various hardware configurations (right). Notably, as seen around 25, larger reductions are observed for systems with lower expected drag. Each experiment repeated 5 times. | 2012.12403 | adtmpc_validation.jpg |
0.421155 | 6e70fab09ac84b44b9be19ea7c8bc638 | DTMPC | 2012.12403 | disturb_dtmpc.jpg |
0.434656 | 15914e57dda64b8bbe6cdb8ddff84848 | TMPC | 2012.12403 | disturb_tmpc.jpg |
0.447383 | 9ac481863c714d35948ebc2cd5dc4f3b | Comparison of DTMPC and ADTMPC control policies: angular speed and the upper bound on C_d. With every cycle, ADTMPC reaches higher speeds due to the decrease in C_d, until reaching steady-state. DTMPC performance is provided as the baseline to ADTMPC. Experiment repeated 3 times. | 2012.12403 | dtmpc_adtmpc_comparison.jpg |
0.510917 | 23ee04a720bb403399bb1ce741f82d96 | A free body diagram of a pendulum. | 2012.12403 | pendulum.jpg |
0.481782 | 003364d4e81c47b8acd2e00ad44f200c | The robust control invariant (RCI) tube Ω_x around the desired trajectory x^*. If the state x begins inside the RCI tube, it remains inside the RCI tube indefinitely for all realizations of uncertainty <cit.>. | 2012.12403 | rci_no_tilde.jpg |
0.541552 | 4a01959bedee42dbad445a37ddb83fd9 | DTMPC | 2012.12403 | tube_dtmpc.jpg |
0.39815 | a34a183850ff45bb9d294bd9e2b69147 | TMPC | 2012.12403 | tube_tmpc.jpg |
0.41957 | 859eea8fcaec469a95669f2325601229 | Flat Plate attachment | 2012.12403 | white_flat_plate.jpg |
0.434593 | 28c048a9d19a426694ef0e1f605d7416 | No attachments | 2012.12403 | white_no_attach.jpg |
0.461522 | 18679f8dca154fc29d03ea2038f2978f | Scoop attachment | 2012.12403 | white_scoop.jpg |
0.46637 | 91a506f3c6ea4fc6ba4a78c4298f583d | Atom-projected electronic band structure and density-of-states (DOS) of monolayer (1L) (a) Tl_2O and (b) PtS_2, respectively. Spin-orbit coupling is included in both cases. | 2012.12408 | fig1.jpg |
0.440781 | 4f1a18b6ec8a49f6a7cabfc7fe498dfc | (a-c) Top views and (d-f) side views of Tl_2O/PtS_2 heterostructure in different lateral stacking configurations. Unit cell is shown as black rectangle for each case. | 2012.12408 | fig2.jpg |
0.498328 | 0ca11ef6f15142a4b3c525f9366a870f | Electronic band structure of Tl_2O/PtS_2 heterostructure (a) without (w/o) SOC and (b) with (w/) SOC, respectively. (c) Zoomed valence band region defining the generalized Rashba-energy E_R and the momentum-offset k_0. (d) Fixed-energy spin-contour plots corresponding to dashed black line (E=-0.30 eV) in (c). | 2012.12408 | fig4.jpg |
0.395384 | 57bffceccd5a4470866a09c5f1e210e3 | Electronic band structure projected on (a) Tl_2O (b) PtS_2. (c) 2D Electron localization function corresponding to the (110)-surface passing through the heterostructure. (d) Total and atom-projected density of states (DOS). | 2012.12408 | fig5.jpg |
0.479996 | ab21fd0805344feaa7e2032df58f049e | Raman spectra of pristine Tl_2O (black) and Tl_2O/PtS_2 heterostructure (red). | 2012.12408 | fig6.jpg |
0.470547 | f2199309643d4c319b02bfab9232ce91 | Effect of strain on the electronic band structure of monolayer Tl_2O/PtS_2 heterostructure. | 2012.12408 | figs1.jpg |
0.486472 | 648cceaf1c804b138df4e194aa872de8 | Atomic orbital projected band structures of the Tl_2O/PtS_2 heterostructure. | 2012.12408 | figs2.jpg |
0.427036 | 96e452bad33e42e0a8b23a49062eb366 | The interpolated cross-correlation function (ICCF) as a function of the time-delay expressed in days in the observer's frame. The calculation is performed for the d11_ mod template and the redshift of z=1.2231. Right panel: The ICCF as a function of the time-delay expressed in days in the observer's frame for three types of interpolation: interpolated continuum light curve (green solid line), interpolated line light curve (blue solid line), and the symmetric interpolation (black solid line). Middle panel: The distribution of the cross-correlation centroids for the symmetric interpolation. Right panel: The distribution of the cross-correlation peaks for the symmetric interpolation. | 2012.12409 | HE0435_iccf.jpg |
0.427416 | f6e22ce457974fdc9b54195608e50257 | The same ICCF analysis is in Fig. <ref> but for the KDP15 template (case f in Table <ref>) and the redshift of z=1.2330. | 2012.12409 | HE0435_iccf_serbian.jpg |
0.41787 | db95cf985b234ff99eddf401076f6c20 | The zDCF as a function of the time delay in the observer's frame. The uncertainties in each population bin were inferred from 5000 Monte-Carlo generated pairs of light curves (using the observed measurement errors). The red vertical line marks the most prominent, maximum-likelihood peak of 646^+63_-57 days, while the green vertical dashed lines mark the uncertainties. | 2012.12409 | zdcf_lag_HE0435.jpg |
0.422588 | c709d79d601047a7937080583bcd7238 | Light curves and the time evolution of emission line properties measured in the rest frame. Points marked with a black circle (obs. 19) were removed from the analysis. | 2012.12409 | zestaw1_v2.jpg |
0.468743 | 9eb20cea6d574af18978c8b52dc6d529 | Selected models of observation 23 from Table <ref>. | 2012.12409 | zestaw_obs23.jpg |
0.475856 | 82d07410cc8044c7910be6a21e2ff1a1 | Upper panel: mean spectrum (red curve) obtained without observation 19 and its decomposition into two Lorentzian components of MgII line (dotted magenta), FeII pseudo-continuum (dotted blue line), and a power law (green dashed line); upper middle panel: rms spectrum and its decomposition; lower middle panel: both mean and rms spectra in absolute calibration; lowest panel: A properly calibrated mean spectrum, with an rms spectrum multiplied by a factor of 23 for better comparison of the two shapes. | 2012.12409 | zestaw_rms_mean.jpg |
0.419215 | 42fd891b0fbc4740a96ea54a4c128a69 | Trained Model Figures and Metrics: Dice score, ROC Curve and, Model Confusion Matrix from left to right respectively. | 2012.12410 | MergedFig.jpg |
0.479511 | 9b7f6ec0d0564d0fb9db7c4a72e1bfd7 | QuickTumorNet learning curve | 2012.12410 | ModelLearningCurve.jpg |
0.3758 | 921b27b675504f13bd971f903fad8a17 | Network Segmentation - High DICE Scores: Three brain tumor classes in the three different planes from our local test set. Green delineations are the radiologist annotations (ground truth) and blue borders are the network's predicted multi-class segmentations correctly classified. | 2012.12410 | PredSample.jpg |
0.365564 | 70fb9f0b83014b7588b57e7aa71d4ab8 | Network Segmentation - Low DICE Scores: (Color coding similar as Figure 5) | 2012.12410 | PredSampleBad.jpg |
0.410575 | d271ebcbc01d4771b3702538a1b97c60 | Spatial Distribution of tumors across patients presented in blue-red false color. Tumor class size distributions across data set. | 2012.12410 | TumorSpatialAndSizeDistribution.jpg |
0.413433 | 8ec0c483ad7a41b386b6e10b9fcb700c | Surface fitting for a deformed surface (top row) with 49 markers' positions determined by a motion capture system – the results by using 4 × 4, 5 × 5 and 6 × 6 control points are given from left to right (bottom row). The black points indicate the measured coordinates of markers, the blue points present the control points obtained by surface fitting and the fitted Bézier surfaces are visualized as the purple grids. | 2012.12411 | Figure4_v2.jpg |
0.430607 | 98ee053d18254151ab7c11a94aa9ebef | The hardware setup of a deformable membrane, which is composed of four inflatable modules. The chamber of each module is mounted with 3 LEDs and 3 LDRs to sense the deformation of each chamber, the signals of which are finally fused to predict the surface shape of the silicone layer. Markers are attached for learning the 3D surface shape by a motion capture system. | 2012.12411 | Picture_surface.jpg |
0.408421 | 2ee64be22da849149bdceeee441d2f02 | Comparison of the mean prediction errors by using different learning models for the soft continuum joint and the soft deformable membrane. LSTM provides the best performance for predicting transformations of the continuum joint, whereas FNN provides the best performance for the deformable membrane by using 5 × 5 control points. Note that the surface error is evaluated as the distances between real and predicted positions of markers. | 2012.12411 | bargraph_error_ML_models_highres.jpg |
0.489807 | 687797d70a0e459faf65ee904028cfa6 | Comparison of the mean distance error for the fitted surface vs. the captured markers, and the predicted surface vs. the captured markers on Bézier surfaces with different number of control points. | 2012.12411 | bargraph_error_controlpoint_resolution_highres.jpg |
0.415998 | 5c1cb298915044cf9e93a651e6062e08 | Overview of our approach for enabling the sensing capability of 3D shape on soft robots by four steps: (1) sensor integration, (2) data collection, (3) shape parameterization and (4) learning the mapping. | 2012.12411 | figure_approach.jpg |
0.408251 | 0ff38893a2ed4544a233e80e7b00795d | Histogram of distance errors between real and predicted positions of markers on the soft deformable membrane when using different numbers of LDRs in each module. | 2012.12411 | histogram_FNN_membrane_error_numLDRs.jpg |
0.41464 | 3ce938c561ba4608abb4cefa346f4fb2 | Overall diagram of the proposed approach. Phase 1, translation models are learnt to translate images from the source domain to the target domains; Phase 2, patches of the source domain are randomly translated to the target domains during training (U-Net image taken from <cit.>). | 2012.12413 | diagram.jpg |
0.494355 | b4990b7a3ccd4c28b644e64956780abe | Overview of point-vortex states at negative, zero and positive inverse temperatures β. Clustering occurs for β<β_c<0, a homogeneous state is found at β=0, and pair condensation occurs for β>β_pc. | 2012.12637 | fig1.jpg |
0.483444 | c6883eac61624d98859f69c653d0a2a6 | PDF of sample mean ⟨ A⟩, over n_sample realizations (independent ergophages) from the passive nonlinear point-vortex model with parameters δ=1, ν/ν_c=0.15. For n_sample=1, the PDF is close to the theoretical prediction for the non-truncated system, and converges to a Gaussian PDF (thin dashed line) as n_sample is increased. | 2012.12637 | fig10.jpg |
0.426528 | 3d36030b15c54d8eb778adf4bdcf4f3a | Dynamics of the QDC in the CAD. Panels from bottom to top: the exciting laser pulse, the occupation of the exciton n_X, the average photon number n, (_̊photon^2) which indicates how close the photonic system is to a pure state, and the fidelity to the Schrödinger-cat state in Eq. (<ref>). Dashed lines: ideal case without phonons and losses. Dashed-dotted lines: without phonons but with losses. Solid lines: with phonons and losses. The temporal range where the fidelity reaches its maximum value in all three considered cases is shaded in gray. | 2012.12767 | GB_dynamics.jpg |
0.441574 | a268c4c5e22a433d8c7e3f49b2ec73b5 | Dynamics of the QDC in the DOD. Panels from bottom to top: External pulses as well as the Stark pulse, exciton occupation n_X, average photon number n and time-dependent fidelity ℱ(t) to the target Schrödinger-cat state in Eq. (<ref>). Dashed lines: ideal case without phonons and losses. Dashed-dotted lines: without phonons but with losses. Solid lines: with phonons and losses. | 2012.12767 | LE_dynamics.jpg |
0.516136 | 5e04bb58d8ad4c93bc234ed7832364e0 | (a) Nonclassicality measure δ at the time of maximum fidelity and (b) maximum fidelity during the time evolution as a function of the pulse area, i.e., the average photon number of the initial state. Black: ideal case, blue: including losses but without phonons, red: with losses and phonons, orange: solution of the initial value problem . | 2012.12767 | area_fidelity_nonclassicality.jpg |
0.475856 | d13820d2da8b418a9acd24372a385c5f | The maximum fidelity to the Schrödinger-cat states in Eqs. (<ref>) and (<ref>) in the case of the DOD (black solid line) and the CAD (red dashed line), respectively, and the nonclassicality measure δ (open rectangles and circles) as a function of the cavity loss rate κ. The cavity quality factor assuming a mode frequency of ħ_C=1.5meV is displayed as a second axis. | 2012.12767 | kappa.jpg |
0.498048 | 4b70ac30b35f4de8914d18372c91ad57 | Wigner functions of the photonic states prepared by the DOD (a) for the ideal case, (b) including losses, and (c) taking both phonons and losses into account. The Wigner functions are calculated at the start time of the decoupling Stark pulse, i.e., at t=75.5ps. | 2012.12767 | plot_Wigner_triple.jpg |
0.559038 | 43db2246da3440558c9134365264ac42 | Wigner functions of the photonic states prepared by the DOD (a) for the ideal case, (b) including losses, and (c) taking both phonons and losses into account. The Wigner functions are calculated at the start time of the decoupling Stark pulse, i.e., at t=75.5ps. | 2012.12767 | plot_Wigner_triple.jpg |
0.398843 | 40450573576344c9beb64ce193eb8fa2 | A Schrödinger cat state appearing in a QDC system. Photons are created by recombination of the QD exciton. The photoemission is controlled such that cats are created on demand. | 2012.12767 | sketch.jpg |
0.437124 | 2fbef45bd99540a7be5cc8539a01db96 | Synthesis results for string editing. Blended semantics outperforms all baselines, except for the execution-guided REPL model, which relies on domain-specific language modifications. | 2012.12964 | SMCChallengeRebuttalTransparentrbtest_curve_evaluations_mod2.jpg |
0.452069 | f9520b5892164d6f9582d09834ee8967 | Schematic overview of the search procedure and representational scheme. We characterize program synthesis as a goal-conditioned search through the space of partial programs (left), and propose a novel representational scheme (blended abstract semantics) to facilitate this search process. Left: a particular trajectory through the space of partial programs, where the goal is to find a program satisfying the target image. Right: three encoding schemes for partial programs, which can each be used as the basis of a code-writing search policy and code-assessing value function. | 2012.12964 | fig1_camera_ready.jpg |
0.480171 | 2c0897681f504127a8b15274b42d1558 | Synthesis results for list processing. Models were trained and tested on programs each containing 2-3 higher order functions with lambdas of depth 3. | 2012.12964 | mainpaper.jpg |
0.447054 | 81c9891f606e48d9bf18decde4f45283 | (a) Example applications of the function. (b) Neural abstract module for . (c) Neural placeholder module encoding a with the context {x=3}. (d) Neural abstract semantic encoding of the partial program with the context {x=3}. | 2012.12964 | new_fig_2_try.jpg |
0.437417 | 1a0530aeb2124ed480ab2ee96112ac6d | Example tower-building constructions. | 2012.12964 | towers_wrap_tmp.jpg |
0.449174 | 1cdee5d03379468faf924146530009be | Example 𝒢(t) calculation. (a) Radiofrequency (RF) pulses in a static gradient of amplitude g. (b) G(t), shifted G(t+s) (red, dashed), and F(t). (c) Corresponding 𝒢(t). G(t) is assumed to be zero outside of t = [0, 2τ]. | 2012.12968 | fig1.jpg |
0.522066 | 1abe844bb10b474ea2f966cbf68ada29 | Example SG-TIETA sequence. (a) Timings: m_j = {1, 3, 1, 2, 1}, τ = 4δ, and δ = 14 μs = 1 dash. π-pulses occur at t_n and direct echoes (blue lines) form at T_n. Magenta line indicates timing behavior: T_n = t_n + h_n, where h_n is the normalized |F(t)|/γ g “height” at t_n, given by h_1 = τ and h_n = 2τ + m_nδ -h_n-1 for n > 1. (b) Direct echo F(t) and other coherence pathways that refocus (red, dash-dot) or do not refocus (gray, dotted). Relative values of h_n are indicated. | 2012.12968 | fig2.jpg |
0.482441 | bf812faf9f804b3786a2cd3d76c44b86 | LLS inversion on Monte Carlo simulated data. (a) Simulated ⟨[ r(t) ·ĝ]^2⟩ (D_0 = 2.15 μm^2/ms) for free diffusion (black) and restricted geometries (see Fig. S1). ∑𝒞_n(t) (a.u.) for Eq. (<ref>) is overlaid, omitting the first two echoes. (b) Echo decays simulated from same color curves in (a) for γ g = 4.093 μm^-1ms^-1. Insets show B with 2W^1/2 for the black curve and A for Δ t(k) = {50, 50, 50, 50, 75, 75, 125, 139} μs. (c) D_inst(t) from the gradient of ⟨[ r(t) ·ĝ]^2⟩ curves in (a) (solid lines) compared to X inverted from decays in (b) with added Gaussian noise (SNR = 25). Err. bars = ± 1 SD from 100 replications. Initial X guesses were D_0 for the first point and D_∞ (dashed lines, {0.9, 1.45, 1.48} μm^2/ms) for remaining points. λ = 2× 10^-6, selected manually. See Eq. (S1) for Γ. | 2012.12968 | fig3.jpg |
0.440515 | 01962700626441ddae6bd5827ae8e395 | SG-TIETA A_p(n) calibration using pure liquids. (a) Observed SG-TIETA decays compared to exp(-bD_0). D_0 was measured (see SM Section IV) in legend order as 2.22 ± 0.01, 1.27 ± 0.01, 0.121 ± 0.002 μm^2/ms. Err. bars =±1 SD for 25, 38, 70 repetitions, respectively. (b) Decay vs. exp(-bD_0) ratio truncated at n = 12, 15, 25. Inset shows cubic spline fits. (c) A_p(n) approximated using adjacent fitted ratios. | 2012.12968 | fig4.jpg |
0.505515 | 5b58130a68fd46c480b8849f97367a4c | Comparison of the direct echo SG-TIETA (blue) and CPMG (orange) attenuation and echo shape for 1-octanol with 2τ = TE = 98 ms. Echo shapes show the real (solid lines) and imaginary (dotted lines) signal normalized by the area under the real signal curve in a 16 μs window. The CPMG echo width decreases with n and stabilizes around n = 3, consistent with the approach to asymptotic behavior described by Hürlimann & Griffin.<cit.> In contrast, the SG-TIETA echo width increases and stabilizes around n = 3, consistent with the direct echo CTP being preferential to the on-resonance signal. See Figs. S7 and S8 for all echo shapes and an exemplar echo decay, respectively. | 2012.12968 | fig5.jpg |
0.427557 | f3463e81e54e4e43b0f5e42ab31cdaad | SG-TIETA decays and inverted X for D6, yeast, and water. (a) Decays analyzed as described in the text. See SM Section II for fitting procedures. Err. bars = ± 1 SD for, in legend order, 38, 3, 4, 25 repetitions truncated at N = 34, 17, 17, 15, respectively. Note erratic early A_p(n) behavior. (b) X solutions. Inversion parameters were identical to Fig. <ref> other than Δ t(k) for D6. Initial guesses of D_0 = 2.22, D_∞ = {0.9, 1.1} μm^2/ms for yeast and D_0 = D_∞ = 0.114 μm^2/ms for D6 were provided. Zoomed plot compares short-time D_inst(t) plotted up to t < 0.08×τ_D. | 2012.12968 | fig6.jpg |
0.50712 | 603a47ea7e6c42919d0571c94d663583 | Band structure of NdAlSi from first-principles calculations. Paramagnetic (PM) state in the PBE approximation, with Nd f states in the core for panel A, and ferromagnetic (FM) state in the PBE+U approximation, with Nd f states in the valence for panel B. Panel C is the paramagnetic PBE+SOC band structure where the positions of the Weyl nodes in the first Brillouin zone are shown in panel D and E for this calculation. Three types of Weyl nodes W_1, W_2 and W_3 are marked. The mirror planes are shown by dotted lines. Panel F is the PBE+SOC band structure calculation for the u-d-d phase of NdAlSi. | 2012.12970 | Fig1-SI.jpg |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.