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
|
|---|---|---|---|---|
0.396175 | 81ab2096990645ec8bd71186818ab950 | T = 2 | 2012.13024 | poe4.jpg |
0.467228 | 632165771eac4353b0df75fd592397d2 | T = 0.999 | 2012.13024 | poe_0.999.jpg |
0.395563 | e1ac45974a0840faaa0c344f169043c7 | T = 0.9999 | 2012.13024 | poe_0.9999.jpg |
0.472131 | 892fefb99400414387c03d9f68cea615 | Synthetic data A1. The red lines denote x → y, and the blue lines denote y → x. For the clarity of the figure, we present the results for N_F=64, 128, and 256. | 2012.13025 | A1_new.jpg |
0.395981 | f47abf4e4257475cbf710cff732bce67 | PCA results: (a) Variations (b) Non-solvable v.s Solvable case | 2012.13026 | PCA.jpg |
0.378287 | a3792c67760a4c33b45d7eb0a7e073d1 | Policy performance comparison. (Case indices are sorted according to random policy performance. A case is considered unsolvable if its y value reaches 10^3.) | 2012.13026 | policy_performance_comparison.jpg |
0.448603 | c3f36021f5404b8ea8caff36d5518881 | The WADD vs ARL for |J|=3,m=729 and a change in the local variance | 2012.13027 | lorden.jpg |
0.520789 | 00d87ce4af5c447e89cca7e8ae30ee21 | The illustration of weights assigned to the target samples with the same pseudo-label. | 2012.13028 | weights.jpg |
0.403123 | 777d32a37c3f4f23985ff06c8920c9c9 | (Color online) Theoretically proposed spin orderings for the quantum S = 1/2 Heisenberg triangular antiferromagnet in the isotropic 2D limit. Structures (a - d) are the coplanar `Y', `UUD', and `V' co-linear spin ordering favored by quantum fluctuations, while (e) represents the competing classically favored non-coplanar 3D cone (umbrella) spin arrangement. Structure (f) represents the alternative high field coplanar spin ordering Ψ (also referred to as a `fan' structure). In zero field, theory predicts that the spins order in a 120 arrangement within the easy plane ; phase transitions from the Y (b) to the up-up-down UUD (c) to the V spin structures (d) occur as the magnetic field is increased <cit.>. | 2012.13030 | PRB_FIG_01plus.jpg |
0.505646 | 4fbff44279bf46deaf37f10c1446c06f | (Color online) Theoretically proposed spin orderings for a quantum S = 1/2 Heisenberg triangular antiferromagnet with interlayer coupling in a 2 layer model. Spins 1 - 3 correspond to layer 1; spins 4 - 6 to layer 2. The orderings are named as follows: (a) the zero field 120 structure, (b) distorted combined Y ,(c) alternating ordered Y, (d) parallel ordered Y, (e) UUD, (f) V, (g) distorted V, (h) V^', (i) staggered V^', (j) ψ (fan) spin structure. Ordering (k) corresponds to the classically expected cone (umbrella) phase. | 2012.13030 | PRB_FIG_02.jpg |
0.447474 | e3a80d68414644ecbaf027dec7c0a8cb | (Color online) Magnetic-field dependence of the specific heat as a function of temperature for H||a. Features corresponding to magnetic-field induced phase transitions are observed at 6, 10, 15, 24, and 33 T in the low temperature limit. The 6 T transition has not been previously reported. | 2012.13030 | PRB_FIG_03.jpg |
0.412907 | c562eb6946a241b5b97aa18fecab8595 | (Color online) Magnetic-field dependence of the thermal conductivity for H||a at temperatures above and below the zero-field magnetic ordering temperature T_N(0) = 3.7 K, revealing field induced phase transitions from the zero field paramagnetic T > T_N(0) and antiferromagnetic T<T_N(0) states to the UUD spin state. | 2012.13030 | PRB_FIG_04.jpg |
0.48355 | 3060de36c60343fbb09f790a0b9bc1c2 | (Color online) Magnetic field dependence of the thermal conductivity for H||a revealing the onset of a new magnetic phase between 6 and 10 tesla below 1 K. | 2012.13030 | PRB_FIG_05.jpg |
0.462457 | d98d7fcfb7114ca5ab7becb6d41ef8d6 | (Color online) Intensities of three magnetic Bragg scattering peaks â (1/3, 1/3, 0), (1/3, 1/3, 1), and (1/3, 1/3, 2) â at 8 T as a function of temperature for H||a. | 2012.13030 | PRB_FIG_06b.jpg |
0.486256 | 40dc2f359c51411b912ab5336e696baf | (Color online) Magnetic-field dependence of the magnetic torque τ at 0.32 K for a Ba_3CoSb_2O_9 sample for H || a (after scaling for small changes in field orientation). A derivative of τ has been taken with respect to magnetic field to remove a background term in the magnetometer response. Phase transitions observed in the specific heat at 6, 10, 15, 24, and 32 T appear here as extrema and breaks in slope, confirming the magnetic nature of these transitions. The dashed line is a guide to the eye. | 2012.13030 | PRB_FIG_07.jpg |
0.499755 | 17da947416ec417487cf8744a3c96877 | (Color online) Phase diagram for H||a (ϕ = 90). Black circles correspond to the locations of phase transitions observed in this work, blue diamonds to earlier, lower field specific heat measurement by Zhou et al. <cit.>, and red squares to NMR measurements by Koutroulakis et al. <cit.> Hollow black circles correspond to a possible additional high field transition more clearly resolved for ϕ <90. | 2012.13030 | PRB_FIG_08.jpg |
0.413596 | 8001b7625c44400981a8765bf899c9c6 | (Color online) Variation of the specific heat of Ba_3CoSb_2O_9, at 0.4 K, with magnetic field strength and orientation for magnetic field orientations near H||a. The horizontal axes represents magnetic field strength and magnetic field orientation ϕ with respect to the c axis (normal to the easy plane). The horizontal plane serves as an (H, ϕ) phase diagram; the dashed lines on the phase diagram represent the predictions of T = 0 semi-classical mean-field theory <cit.>. The solid dots on the phase diagram represent locations of experimentally observed magnetic-field-induced phase transitions; dashed vertical lines leading from the dots mark the corresponding features in the data. | 2012.13030 | PRB_FIG_09.jpg |
0.415519 | f74e78c7057640fa895c67c397e86305 | (Color online) Variation of the specific heat of Ba_3CoSb_2O_9 with magnetic field strength and orientation for magnetic field orientations near H||c (ϕ = 0) at T = 1 K. The horizontal axes represent magnetic field intensity and magnetic field orientation ϕ with respect to the c axis. The horizontal plane serves as an (H, ϕ) phase diagram; the dashed lines on the phase diagram represent the predictions of T = 0 semi-classical mean-field theory <cit.>. The solid dots on the phase diagram represent locations of experimentally observed magnetic-field-induced phase transitions; dashed vertical lines leading from the dots mark corresponding features in the data. | 2012.13030 | PRB_FIG_10.jpg |
0.544036 | e27f20ee681f4cddb05e3c49a8d41915 | (Color online) Magnetic-field dependence of C_p/T for H||c from 9 T — just below the UUD phase transition at 12 T — up to the saturation field H_s = 33 T. Features corresponding to magnetic-field induced phase transitions are observed at 12, 20, 29, and 33 T at 0.41 K. Lower-field data was not collected for this field orientation. | 2012.13030 | PRB_FIG_11.jpg |
0.57775 | b1c9f98e97394d1db244b1b217295148 | (Color online) Magnetic-field dependence of the torque τ at 0.32 K for H||c. Phase transitions observed in the specific heat for this field orientation at 12, 21, and 33 T are manifest here as extrema and breaks in slope, confirming the magnetic nature of these transitions. | 2012.13030 | PRB_FIG_12.jpg |
0.491319 | b8611bb933cc4bbc997dff14abda615c | (Color online) Phase diagram for H||c (ϕ = 0). Black circles correspond the locations of phase transitions observed in this work, and solid (hollow) red circles represent transitions definitively (tentatively) identified by NMR measurements <cit.> at 1.89 K between 14.5 and 30 T. | 2012.13030 | PRB_FIG_13b.jpg |
0.499569 | faf57090cef845aaa4db60319581603d | Scaling in different dimensions | 2012.13032 | Fractaldimensionexamplebw.jpg |
0.511456 | ccff899ba37e404c9b1769324d8e723d | A non uniform mesh | 2012.13032 | Snow_Mesh_Example.jpg |
0.469041 | e9cc4cfdf6904af58026a507a3261826 | A render of the hopper system studied in this work. | 2012.13032 | hopper_crop.jpg |
0.430885 | 57acd08f7e1843c296a61007775f9f4a | Estimated mean first passage time computed from <ref> compared to a Monte Carlo estimate. The blue dashed line and shaded region are the mean and standard deviation of the steps to failure for 2500 Monte Carlo rollouts. | 2012.13032 | mdim_mfpt.jpg |
0.387678 | ec8d518eb1114a09ba7f02c9c8b1d0a2 | View of the first 3 principle components of the mesh for a fractal noise policy. | 2012.13032 | mdim_pca_side.jpg |
0.408928 | 7c111e1e8f124269a1538964ec864812 | Mesh sizes for the top performing standard noise and fractal noise agents. | 2012.13032 | mesh_size_cmp.jpg |
0.500809 | 2092ca9ca1d54a639aafa747513b817f | Cumulative sum of probability mass excluding the failure state. We take the sum of each column of T, and sort it in descending order, then report the cumulative sum of probability. Each point on the curve tells us that x% of states make up y% of all state transitions. | 2012.13032 | pmass.jpg |
0.444421 | 1318537de8f24f52af5e3e1923628383 | Visualization of the stochastic transition matrix for the top performing fractal noise agent. All non zero values are shown with equal size and coloration. Recall that each entry in T_ij tell us the probability of transitioning to state j after one step if we start in state i. | 2012.13032 | sT2.jpg |
0.461476 | 3fe60fe8df8545768f7a304b8f0a8cc2 | On the left: episode length during training, i.e., how many steps the agent takes to reach the goal. On the right: log plot of the cumulative constraint violations during training, i.e., the number of times the agent has bumped into a wall. Solid marks correspond to the average value over N=1000 independent trials, while shaded regions are the average value ±σ/√(N), with σ the sample deviation. Notice that the learning curves (left) for both algorithms are similar. However, the Assured agents learn to avoid walls, why the other agents do not (right). | 2012.13036 | c3_results2.jpg |
0.468753 | 3a3225e9a9b144b5a2cec2d434b21cd7 | Example grid-world. The agent starts at the green square and if he reaches the red “goal" achieves a reward of +10. Moving around has null reward and bumping into walls is unsafe. With discounting (γ<1), the optimal trajectory is the one that goes through the corridor (shown in blue). During learning, the corridor is potentially unsafe since many actions lead to a wall. As a result, some agents converge to sub-optimal trajectories that avoid the corridor (as the one shown in brown). | 2012.13036 | grid_with_policy.jpg |
0.481133 | f0185155b35c409f8fd742655ad292ce | Baselines up to 200,000 episodes normalized to maximum reward obtained by SPOTTER. | 2012.13037 | learning_curve_all_baselines_200K_600dpi_F.jpg |
0.46644 | e7157d863489477ca0465c3ceb496759 | Baselines up to 2,000,000 episodes normalized to maximum reward obtained by SPOTTER. In the long run, the baselines perform better than SPOTTER, however, they take much longer and do not learn transferable knowledge as effectively as SPOTTER. | 2012.13037 | learning_curve_all_baselines_2M_600dpi_F.jpg |
0.472077 | cbd145d947dd41cfb5713bb9632689d8 | Baselines up to 500,000 episodes normalized to maximum reward obtained by SPOTTER. | 2012.13037 | learning_curve_all_baselines_500K_600dpi_F.jpg |
0.389738 | eaf6ab2b47d34096ac18e8489af58ca6 | By episode 30,000, almost all operators have been dominated (replaced by superior operators), and almost all non-dominated operators were created late in the run. | 2012.13037 | precond_refinement_hist.jpg |
0.454128 | 735354b1d2724a1381bfb7d55e01d95b | The number of unique sets of preconditions discovered for which average value is above threshold increases as SPOTTER is allowed to continue exploring. | 2012.13037 | precond_refinement_num_ops.jpg |
0.417006 | eb2586b16a8c4738acb49fe936727eeb | Generation of operators with decreasing numbers of preconditions proceeds in “waves” as the value estimates of additional MDP states converge. | 2012.13037 | precond_refinement_scatter.jpg |
0.390527 | 5e553e35942a4f9aa2629d99b1691812 | The agent's (red triangle) goal is to unlock the door (yellow square). SPOTTER can learn how to move the blue ball out of the way. The learned representation is symbolic and can be used to plan (without additional learning) to achieve different goals like reaching the green square. | 2012.13037 | screenshot.jpg |
0.472171 | 33c20845c9944cf8ba540b9456770b96 | The red solid line represents C_l_1, and the black dots are -(ρ W_2) for the density matrix in Eq. (<ref>), where p = 0.1, ϕ_1 = 0.2, and ϕ_2 = 0.3. The negative values of -(ρ W_2) are ignored for the ineffectivity of W_2 under such situations. | 2012.13038 | Fig1.jpg |
0.49424 | 55eb929a20534a7ebbfbc7e3028e2355 | The Black Box implements an unitary U_φ=exp(-iH φ) on an input quantum state ρ_in with its output state being ρ_out=U_φρ_in U_φ^†. Suppose that the full information of non-degenerate Hamiltonian H is already known. The quantum metrology task is to estimate the unknown parameter φ by measuring the output state ρ_out. | 2012.13038 | Fig2.jpg |
0.479457 | 7b8904f60c0d4e1db54b8bf8a99bbad4 | Simplicial representation of the Lotka-Volterra model. The representation contains higher-order simplices which capture the interactions between the components of the mathematical model. The vertices labelled `Prey', `Prey growth', `Predation', and `Oscillatory population' span a 3-simplex; the vertices labelled `Predator', `Predator growth', `Predation', and `Oscillatory population' span a 3-simplex; and the vertices labelled `Predator', `Predator growth', `Predator death', and `Oscillatory population' span a 3-simplex. | 2012.13039 | MainFig1.jpg |
0.534407 | 3a647613dadf4f9d8da0ca7f40c31cd4 | Bisubstrate reactions represented schematically and as associated labelled simplicial complexes. (a) Ordered sequential bisubstrate reaction. (b) Random sequential bisubstrate reaction. (c) Ping-pong bisubstrate reaction. | 2012.13039 | MainFig2.jpg |
0.434681 | f72396b9378c43a486fc856b82ec0097 | Ordered set of components for the four Turing-pattern and five positional-information models. | 2012.13039 | MainFig3.jpg |
0.404863 | 0cdcc44dc2de4b74bffc178eba306675 | Simplicial-complex representations and corresponding persistence barcodes. Positional-information annihilation model: (a) 1-skeleton of the 4-dimensional simplicial representation; (b) Persistence barcode. Turing-pattern activator-inhibitor model: (c) 1-skeleton of the 5-dimensional simplicial representation; (d) Persistence barcode. Note that the H_n labels on the persistence barcodes refer to the n-dimensional homology classes, and the arrows at the right endpoints of certain persistence intervals indicate that the right endpoints are + ∞, corresponding to homology classes that are not annihilated in the filtration. | 2012.13039 | MainFig4.jpg |
0.464659 | 47d6a013e5f641b1a5bdc7417a237673 | Model components considered as equivalent. | 2012.13039 | MainFig5.jpg |
0.481949 | 047037ffb17842ac85ad103393d7169c | (a) Schematic crystal structure of the VOI_2 conventional cell (red = V; green = O; purple = I). (b) Projected band structures of the undistorted monolayer VOI_2 for the NM state. The Fermi level is shown with dashed lines. The weight of each vanadium orbital is represented by the size of the circles. The coordinates of the high symmetry points in the plane Brillouin zone (BZ) are Γ = (0, 0, 0), X = (0.5, 0, 0), S = (0.5, 0.5, 0), Y = (0, 0.5, 0). (c) Fermi surface. (d) The energy splitting of V's 3d orbitals with the d^1 configuration. | 2012.13040 | Fig1.jpg |
0.408421 | df0746a37c874701a9633ea24294f02f | (a-c) The phonon spectrum of VOI_2 for the undistorted, DIM-I, and FE-II phases. Here, the 4×4×1 and 4×2×1 supercell was selected for the undistorted and DIM-I/FE-II phases in our calculations with the non-magnetism state, respectively. (a) Undistorted, (b) Dimmerized-I, and (c) FE-II cases. (d) Sketch of the possible structural patterns studied here. (e) The group theory analysis for the monolayer VOI_2. | 2012.13040 | Fig2.jpg |
0.466571 | 89bd42fb90ca422193831375d5c206f9 | (a-b) Schematics of the empty p-d hybridizations in the ferroelectric monolayer VOI_2 at the Γ point. Here, the colored solid dots denote V (red) and O (blue). (a) V(3d_3z^2-r^2)-O(2p_z) hybrid state. (b) V(3d_xz/yz)-O(2p_x) hybrid state. (c) Orbital-resolved density of V 3d(d_xz/yz and d_3z^2-r^2) and O 2p (p_x and p_z) states in the FE-II phase of monolayer VOI_2. “Centrosym” refers to the atoms in the centrosymmetric position while “displaced” indicates the atoms present ferroelectric displacements along the V-O bond ([100]). (d) Ferroelectric distortion mode (Γ^4-) obtained from the undistorted phase of monolayer VOI_2 vs. doping. | 2012.13040 | Fig3.jpg |
0.405017 | 5fd1862a5d434e88951a5c43338c6357 | (a) Band structure of the FE-II monolayer VOI_2 using the NM state. (b) The electronic density of the valence band near the Fermi level (-0.5 to 0 eV ) in the bc plane, showing clear d_ xy characteristics. (c) Electron localization function of the FE-II monolayer VOI_2 in the ab plane. | 2012.13040 | Fig4.jpg |
0.451255 | a489bce82e254a2da303ee2b31a9f5af | (a) Wannier function of the V 3d_xy orbital, indicating the alternating ddσ-bonding and ddσ^*-antibonding states along the b-axis. (b) Different hoppings along the chain used in our DMRG calculations (t_1=-0.647 eV, t_2=0.067 eV and t_3=0.009 eV). All the hopping amplitudes were obtained from the DFT calculations via Wannier functions. The spin arrows represent short-range antiferromagnetic-antiferromagnetic coupling along the chain. (c) S(r)=⟨ S_i · S_j⟩ (with r=|i-j|) and (d) S(q_y) for different values of U/W. Here, we used a 36-sites cluster chain with two nearest-neighbors and one next-nearest-neighbor hoppings. (d) Spin structure factor at the six values of U/W of panel (c). | 2012.13040 | Fig5.jpg |
0.391621 | 0027216dc5124ef7a1f51035179a7db7 | The right and left moving electromagnetic waves in the superlattice layers. | 2012.13041 | Fig1.jpg |
0.466262 | f466f308f27944a9b231537be09505e7 | Reflection and absorption factor for a single layer when the incidence angle is θ_i=π/4. In (a), the reflection coefficient as function of the frequency, for different thicknesses of the silver layer. In (b), the absorption factor a grows as the layers' thicknesses d_c increase. The absorption factor vanishes at ω=0 and ω=ω_p, and its maximum shifts to higher frequencies as d_c grows. | 2012.13041 | Fig10.jpg |
0.426872 | 459a796a3420460f99eb2ba098f6ae6b | Reflection and absorption factor for a superlattice with n=12, and two values of the incidence angle. The incidence angle at the left is π/6, and π/4 at the right. The bandwidths and the absorption factor grow with the incidence angle θ_i and the conducting layer thicknesses d_c. | 2012.13041 | Fig11.jpg |
0.438933 | 5bc3fae69ef446a49082b634543382e7 | Transmission and reflection coefficients for a superlattice with n=8, d_a=50nm, d_c=10nm, and incidence angle θ_i=π/4. In the left, upper part, we have also the predicted resonant frequencies from the dispersion relation. In the right the absorption factor. | 2012.13041 | Fig12.jpg |
0.418127 | 8c2a65dcc4c84afda425cd5b27140cb7 | Constant-phase and constant-amplitude planes for the electromagnetic fields propagating in the conductor media. | 2012.13041 | Fig2.jpg |
0.42937 | 6b7cdf83ce5343d1a7266559cccb245e | Determinant of the transfer matrix M'_c in (<ref>), and the real angle as functions of frequency, conductor-layer width and incidence angle. (a) The determinant of the unit-cell transfer matrix, M'_c, decreases as the conductor layer width d_c increases. (b) The cosine of the real angle ψ. We show here that even though that for ω>ω_p it is close to 1, as is well known, for ω<ω_p is close to zero, except for θ_i=0. | 2012.13041 | Fig3ab.jpg |
0.389988 | 712b3c71f6e641aba09646bc5b16b3a2 | Transmission coefficient as a function of ω and d_a, for a metallic superlattice with n=12, d_c=15nm and θ_i=π/3. In (a) the bands move to lower frequencies as d_a grows. In (b) we see more clearly the resonant transmission for a fix value of the conductor layer width d_c. In the upper part of (b), bands and resonances predicted by the resonant dispersion relation (<ref>). | 2012.13041 | Fig4.jpg |
0.421239 | 864b77f7a10d40b9aa99abaf2ccd7045 | Transmission coefficient as a function of ω and d_c, for a metallic superlattice with n=11, d_a=150nm and θ_i=π/12. In (a), the band width diminishes as the conductor layer width d_c increases. This effect is shown in (b) for three values of d_c, indicated with white arrow in (a). | 2012.13041 | Fig5.jpg |
0.42244 | ee77f5e4a24e45ce9e8e9bba0a36f933 | Transmission coefficient as a function of ω and θ_i, for a metallic superlattice with n=12, d_a=150nm and d_c=5nm. In (b), the transmission coefficient for θ_i=0, π/4 and pi/2. These graphs correspond to those in (a) indicated with white arrows. | 2012.13041 | Fig6.jpg |
0.475732 | 72414f69ed364585ad98e87bf67276e8 | The reflection coefficient R_n as a functions of the wavelength λ. In (a) for a single layer of silver. In (b) for a metallic superlattice with n=10. At the plasma wavelength λ_p, we see a characteristic Fano-like resonance, due to the interference between propagating and evanescent modes. | 2012.13041 | Fig7.jpg |
0.379532 | 4dc367bda7844fedb79de71c931d3410 | Transmission coefficients as functions of d_a and θ_i plotted here to compare with those in figures <ref> (a) and <ref> (b). The agreement is perfect. | 2012.13041 | Fig8.jpg |
0.45394 | 586cd2fa0c844013b4b2b814977de656 | The effect of phase φ in the transmission coefficients. The graph in (b) is a zoom of the graph in (a). The effect of the phase φ is negligible, and null for frequencies above ω_p. | 2012.13041 | Fig9b.jpg |
0.524105 | 415b4f1c74ee48049f030949502ba19c | Space time evolution of a Gaussian wave packet through metallic superlattices with n= 1, 2, and 10, in the left, middle and right, respectively. In the upper panels we show both the transmission coefficients and the Gaussian wave packets at the frequencies at which they were defined at z_o and t=0. In the lower panel, we have the reflected and transmitted wave packets at t=2 |z_o|/c+τ, where τ the tunneling time of the wave packet components of frequency ω_o. The green and red Gaussian curves, define the positions at which the reflected and transmitted packets should be if τ is truly the tunneling and reflecting time. Notice that the reshaping of the transmitted and reflected packets are strongly determined by the transmission coefficients in the frequencies domain where the packets are defined, at a resonance in the middle and at a gap in the right. | 2012.13041 | GrafsDifSLn.jpg |
0.43553 | f77939673c164bb59a7c6ae71a49c9fd | The structure of our multi-model network | 2012.13042 | model.jpg |
0.442382 | b565b4bb79414ccdb939a8d47d9b0dc9 | Pre-event resource allocation results with different PV penetration levels. | 2012.13043 | Large_diff_PVs_v2.jpg |
0.515573 | c4b82998fbce4d4e976d460407a9c1aa | Load served percentage comparison of proposed model with various PV penetration level and base model. | 2012.13043 | Served_load_diff_PVs.jpg |
0.495737 | e68c998c78bc45c2bb8cb02d383c9175 | A single line diagram of an example network with one damaged line. | 2012.13043 | fig2_1.jpg |
0.430711 | 17cf2620004f4500bf5dee7e389adddb | A virtual network created for the example network in Figure <ref>. | 2012.13043 | fig2_2.jpg |
0.409324 | 227d73dbce44454d96c1a4e3690278a0 | A crew assignment example with 2 depots and 5 crews. | 2012.13043 | fig2_3.jpg |
0.538258 | de39757385264dbd98ca9871c49e94e5 | Comparison between base model and proposed method. | 2012.13043 | fig2_52.jpg |
0.426773 | 2881b2180bb24a4b8e73745cdc77d4a5 | Aggregated damaged areas. | 2012.13043 | fig2_6.jpg |
0.434321 | 85646f5df9b1404a94801f57ddc18b9a | The convergence metric comparison with and without soft-start solutions. | 2012.13043 | fig2_7.jpg |
0.428977 | 6000066bd725418389800b008aef9919 | Resource allocation of large-system with the proposed model. | 2012.13043 | fig_large_results_v2.jpg |
0.412825 | d46a01b5a6eb49469fe1567a50e27074 | The proposed two-stage stochastic pre-event model. | 2012.13043 | framework.jpg |
0.429554 | 1a7fc047ebe4439c879bd68b8448934f | Our SE simulations of the planar state in Hookean sheets with various aspect ratios, L/W, suggest the existence of a parameter regime (i) where the stress may be purely tensile ans the planar state is thus stable (LW < 2d^*); (ii) where transverse compression exists around the middle of the sheet (2d^* < LW < 2d_m^*); and (iii) where transverse compression exists in two zones (LW > 2d_m^*). For d^*, the numerical values extracted from our SE simulations is d^* ≈ 0.625, whereas for d_m^* the extracted value is d_m^* ≈ 0.8 – 1.0 (the uncertainty is a consequence of a very shallow local maximum of (x,0).) | 2012.13046 | length_threshold_all.jpg |
0.431088 | 2cafdf1c68bb491297932e4cef887d9c | ProofWriter iteratively generates 1-step implications and their proofs, and adds implications back into into the context for deeper reasoning. The stepwise proof fragments are assembled into full proofs of N-hop conclusions. | 2012.13048 | 1hop-proofs.jpg |
0.604983 | b14b4ac7409a418eb4f6a5dbd4418389 | [Task 1] All-At-Once proofs can be verified by checking each step as a separate QA query. | 2012.13048 | architecture.jpg |
0.437951 | dd84aed4a96944aa921a4b64994f2b5d | [Task 1] Comparison of the All-At-Once vs. Iterative ProofWriter models, trained on D3 and tested on D5. While scores are mostly similar throughout, the iterative model generalizes substantially better to generate proofs of depths unseen during training (red box). | 2012.13048 | comparison-ood.jpg |
0.460395 | 6bbf2f9b7cfc4595b83a0c032f9d1281 | An example proof that includes a negated (negation-as-failure) fact. | 2012.13048 | naf-proof.jpg |
0.528848 | 5284678c6f5c4762b71d8d6333efd286 | Given facts, rules, and a question all expressed in natural language, ProofWriter answers the question and generates a proof of the answer. | 2012.13048 | proof.jpg |
0.426516 | 52bc8d82f2244425a19d74c00ec6d774 | (A) While the compression level in the wrinkled state vanishes asymptotically as ϵ→ 0, the transversely-compressed zones gets somewhat closer to the clamped edges. We plot here the extent, x^*(ϵ), of the transversely-tensile zone next to each of the clamped edges (i.e. the distance from the clamped edge at which (x,y=0) changes sign for given = 0.01). The red circle shows the asymptotic value, x^*, extracted from the TFT simulations in Sec. <ref> of a sheet with no bending resistance, and the dashed horizontal line indicates the corresponding length in the planar stress. Inset: x^*(ϵ) converges to the TFT value x^* with a residual ∼ϵ^1/3. | 2012.13051 | PaperII_fig3_dist2clamp_withlog_colored_xstar.jpg |
0.401663 | a31010362cf0434fa2703b513885aca2 | annotator 1 | 2012.13052 | cifar_pairwise_anno_1.jpg |
0.432297 | ea515020ad7541529c4c70450ca79459 | annotator 2 | 2012.13052 | cifar_pairwise_anno_2.jpg |
0.453628 | 80a4136f4d084aea8f32d3f4f048357c | annotator 3 | 2012.13052 | cifar_pairwise_anno_3.jpg |
0.449524 | 998fc398770e49ff89f7e3f267b88dae | common noise | 2012.13052 | cifar_pairwise_com.jpg |
0.453099 | ccbd1aecd89e4f9eac033bd8d8091c6e | annotator 1 | 2012.13052 | cifar_random_anno_1.jpg |
0.475657 | 0628aa21c3894169a6f6e99f4adffb76 | annotator 2 | 2012.13052 | cifar_random_anno_2.jpg |
0.452285 | 2c5d0102da5c4032af5970cfb62724c5 | annotator 3 | 2012.13052 | cifar_random_anno_3.jpg |
0.460115 | 4f544c6ce04d4bdc826407539d01668f | common noise | 2012.13052 | cifar_random_com.jpg |
0.483214 | 636ba74e4aff464093740bc7b0d8286a | Common noise model. | 2012.13052 | com_graph.jpg |
0.483214 | a63c5a29d75b4a04bfa8718012d83aa6 | DS model. | 2012.13052 | ds_graph.jpg |
0.474252 | d4518a76be814dd5be1840f3761f2189 | Overview of our framework for classification with 3 classes and R annotators. | 2012.13052 | framework.jpg |
0.386579 | 14aae977f1594cb99f1cfe3a4ca7efc7 | LabelMe | 2012.13052 | labelme_confusion.jpg |
0.378879 | ae7ac135610143c8bf9e001ebb04bf62 | Music | 2012.13052 | music_confusion.jpg |
0.486634 | 37e740fce6dd41609a02ce03cc13c8e7 | annotator 1 | 2012.13052 | pairwise_anno_1.jpg |
0.518163 | 08097d9853e84a21a4f29c27dc5a0c3c | annotator 2 | 2012.13052 | pairwise_anno_2.jpg |
0.476013 | f83e0aa260b347108cbb8c39f2f4178d | annotator 3 | 2012.13052 | pairwise_anno_3.jpg |
0.464449 | 61f1c783911c423481e59d8fe402580a | common noise | 2012.13052 | pairwise_com.jpg |
0.497372 | f84ce611e0514aed8419e88fb6b12f63 | annotator 1 | 2012.13052 | random_anno_1.jpg |
0.485743 | 923449ccb9e241e3b75aadab3002efc1 | annotator 2 | 2012.13052 | random_anno_2.jpg |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.