Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
43
datasets available to search
ShareScore release 0.9.0
Dataset results
43 results for “quantization”
Dataset for Min-Deviation-Flow in Bi-directed Graphs for T-Mesh Quantization
<p>Dataset for the paper <a href="https://www.algohex.eu/publications/bimdf-quantization/"><strong>Min-Deviation-Flow in Bi-directed Graphs for T-Mesh Quantization</strong></a>, <em>Martin Heistermann, Jethro Warnett, David Bommes</em>, ACM Transactions on Graphics, Volume 42, Issue 4 (2023), <a href="https://doi.org/10.1145/3592437">DOI 10.1145/3592437</a></p> <p>Both archive files (.zip and .tar.zst) have identical contents.</p> <p>Contents:</p> <ul> <li>Results of running various algorithmus variants on the "300" dataset <ul> <li>Intermediate data (e.g., patch decompositions)</li> <li>Final meshes</li> <li>Logfiles</li> <li>Metadata in json format (e.g., utilized settings, achieved energies, detailed runtimes)</li> </ul> </li> <li>Meshes used in all figures</li> </ul> <p> </p>
Robustness assessment of a C++ implementation of a quantized (int8) version of the LeNet-5 convolutional neural network
<p>The architecture of the LeNet-5 convolutional neural network (CNN) was defined by LeCun in its paper "Gradient-based learning applied to document recognition" (<a href="https://ieeexplore.ieee.org/document/726791">https://ieeexplore.ieee.org/document/726791</a>) to classify images of hand written digits (MNIST dataset).</p><p>This architecture has been customized to use Rectified Linear Unit (ReLU) as activation functions instead of Sigmoid, and 8-bit integers for weights and activations instead of floating-point.</p><p>It consists of the following layers:</p><ul><li><strong>conv1</strong>: Convolution 2D, 1 input channel (28x28), 3 output channels (28x28), kernel size 5, stride 1, padding 2.</li><li><strong>relu1</strong>: Rectified Linear Unit (3@28x28).</li><li><strong>max1</strong>: Subsampling buy max pooling (3@14x14).</li><li><strong>conv2</strong>: Convolution 2D, 3 input channels (14x14), 6 output channels (14x14), kernel size 5, stride 1, padding 2.</li><li><i><strong>relu2</strong></i>: Rectified Linear Unit (6@14x14).</li><li>max2: Subsampling buy max pooling (6@7x7).</li><li><i><strong>fc1</strong></i>: Fully connected (294, 147)</li><li><i><strong>fc2</strong></i>: Fully connected (147, 10)</li></ul><p>The fault hypotheses for this work include the occurrence of:</p><ul><li><strong>BF</strong>: single, double-adjacent and triple-adjacent bit-flip faults</li><li><strong>S0</strong>: single, double-adjacent and triple-adjacent stuck-at-0 faults</li><li><strong>S1</strong>: single, double-adjacent and triple-adjacent stuck-at-1 faults</li></ul><p>In the memory cells containing all the parameters of the CNN: </p><ul><li><strong>w</strong>: weights (int8)</li><li><strong>zw</strong>: zero point of the weights (int8)</li><li><strong>b</strong>: biases (int32)</li><li><strong>z</strong>: zero point (int8)</li><li><strong>m</strong>: m (int32)</li></ul><p>Images 200 to 249 from the MNIST dataset have been used as workload.</p><p>This dataset contains the raw data obtained from running exhaustive fault injection campaigns for all considered fault models, targeting all considered locations and for all the images in the workload.</p><p>In addition, the raw data have been lightly processed to obtain global data related to the particular bits and parameters affected by the faults, and the obtained failure modes.</p><h3>Files information</h3><ul><li><i>golden_run.csv</i>: Prediction obtained for all the images considered in the workload in the absence of faults (Golden Run). This is intended to act as oracle to determine the impact of injected faults. </li><li><i>single_faults/bit_flip</i> folder: Prediction obtained for all the images considered in the workload in presence of single bit-flip faults. There is one file for each parameter of each layer.</li><li><i>single_faults/stuck_at_0</i> folder: Prediction obtained for all the images considered in the workload in presence of single stuck-at-0 faults. There is one file for each parameter of each layer.</li><li><i>single_faults/stuck_at_1</i> folder: Prediction obtained for all the images considered in the workload in presence of single stuck-at-1 faults. There is one file for each parameter of each layer.</li><li><i>double_adjacent_faults/bit_flip</i> folder: Prediction obtained for all the images considered in the workload in presence of double adjacent bit-flip faults. There is one file for each parameter of each layer.</li><li><i>double_adjacent_faults/stuck_at_0</i> folder: Prediction obtained for all the images considered in the workload in presence of double adjacent stuck-at-0 faults. There is one file for each parameter of each layer.</li><li><i>double_adjacent_faults/stuck_at_1</i> folder: Prediction obtained for all the images considered in the workload in presence of double adjacent stuck-at-1 faults. There is one file for each parameter of each layer.</li><li><i>triple_adjacent_faults/bit_flip</i> folder: Prediction obtained for all the images considered in the workload in presence of triple adjacent bit-flip faults. There is one file for each parameter of each layer.</li><li><i>triple_adjacent_faults/stuck_at_0</i> folder: Prediction obtained for all the images considered in the workload in presence of triple adjacent stuck-at-0 faults. There is one file for each parameter of each layer.</li><li><i>triple_adjacent_faults/stuck_at_1</i> folder: Prediction obtained for all the images considered in the workload in presence of triple adjacent stuck-at-1 faults. There is one file for each parameter of each layer.</li></ul><h3>Methodology information</h3><p>First, the CNN was used to classify all the images of the workload in the absence of faults to get a reference to determine the impact of faults. This is golden_run.csv file.</p><p>After that, one fault injection experiment was executed for each bit of each element of each parameter of the CNN.</p><p>Each experiment consisted in:</p><ul><li>Affecting the bits (inverting it in case of bit-flip faults, setting it to 0 or 1 in case of stuck-at-0 or atuck-at-1 faults) identified by the mask.</li><li>Classifying all the images of the workload in the presence of this fault. The obtained output was stored in a given .csv file.</li><li>Removing the fault from the CNN by restoring the affected bits to its previous value.</li></ul><h3>List of variables (Name : Description (Possible values))</h3><ul><li><strong>IMGID</strong>: Integer number identifying the considered image (200-249).</li><li><strong>TENSORID</strong>: Integer number identiying the parameter affected by the fault (0 - No fault, 1 - conv1.w, 2 - conv1.zw, 3 - conv1.m, 4 - conv1.b, 5 - conv1.z, 6 - conv2.w, 7 - conv2.zw, 8 - conv2.m, 9 - conv2.b, 10 - conv2.z, 11 - fc1.w, 12 - fc1.zw, 13 - fc1.m, 14 - fc.b, 15 - fc1.z, 16 - fc2.w, 17 - fc2.zw, 18 - fc2.m, 19 - fc2.b, 20 - fc2.z)</li><li><strong>ELEMID</strong>: Integer number identiying the element of the parameter affected by the fault (-1 - No fault, [0-2] - {conv1.b, conv1.m, conv1.zw}, [0-74] - conv1.w, 0 - conv1.z, [0-5] - {conv2.b, conv2.m, conv2.zw}, [0-149] - conv2.w, 0 - {conv1.z, conv2.z, fc1.z, fc2.z}, [0-146] - {fc1.b, fc1.m, fc1.zw}, [0-43217] - fc1.w, [0-9] - {fc2.b, fc2.m, fc2.zw}, [0-1469] - fc2.w)</li><li><strong>MASK</strong>: 8-digit hexadecimal number identifying those bits affected by the fault ([00000000 - No fault, FFFFFFFF - all 32 bits faulty])</li><li><strong>FAULT</strong>: String identiying the type of fault (NF - No fault, BF - bit-flip, S0 - Stuck-at-0, S1 - Stuck-at-1)</li><li><strong>OUTPUT</strong>: 10 integer numbers provided by the CNN as output after processing the image. The highest value identifies the selected category for classification.</li><li><strong>SOFTMAX</strong>: 10 decimal numbers obtained after applying the softmax function to the provided output. They represent the probability of the image of belonging to the corresponding category for classification.</li><li><strong>PRED</strong>: Integer number representing the category predicted for the processed image.</li><li><strong>LABEL</strong>: integer number representing the actual category for the processed image.</li></ul>
Quantized versions of fastText embeddings for 158 languages
<p>These are compressed (quantized) versions of word embeddings for 158 languages originally available from https://fasttext.cc/docs/en/crawl-vectors.html . Following steps were performed to reduce file size:</p> <ol> <li>the output matrix is discarded</li> <li>the quantization is performed with parameters "-qnorm -dsub 1"</li> </ol> <p>The average cosine similarity between original and quantized vectors for frequent words is 0.99. The file size is 4-6 times smaller.</p>
Figure 10. Conceptual diagram illustrating vector quantization codeword formation-Development an Automatic Speech to Facial Animation Conversion for Improve Deaf Lives
<p>After the conversion process converts the analogue speech signal into a series words,<br> then converting recognized word to facial animation based on VRML.</p>
Quantized thermal and spin transports of dirty planar topological superconductors
<p>The module contains the code and data corresponding to the paper arXiv:2308.16908, titled 'Quantized thermal and spin transports of dirty planar topological superconductors'.</p>
Interaction between unidirectional quantized vortex rings
<p>All data included in a paper on "Interactions between unidirectional quantized vortex rings". This paper is currently being prepared for submission.</p>
Data for "Quantized critical supercurrent in SrTiO3-based quantum point contacts"
<p>Spreadsheets containing raw data, axis and trace labels necessary to reproduce all figures in the submitted manuscript "Quantized critical supercurrent in SrTiO<sub>3</sub>-based quantum point contacts".</p> <p> </p>
Quantized Thermal Hall Conductance and the Topological Phase Diagram of a Superconducting Bismuth Bilayer
<p>Data for the topological phase diagrams and thermal Hall effect for the paper "Quantized Thermal Hall Conductance and the Topological Phase Diagram of a Superconducting Bismuth Bilayer". The thermal Hall conductance data is normalised as in figure 4 of the paper.</p> <p>Parameters in the Chern number file names are {t,Delta,Alpha, tp}. For the thermal conductance data as. function of temperature the Chern number is given. All over information can be found in the paper.</p> <p><a href="https://doi.org/10.48550/arXiv.2308.01021">arXiv:2308.01021</a></p> <p> </p>
All-atom models of SU10 and P68 viral genomes. Supplementary to the "Are kuravirus capsid diameters quantized? The first all-atom genome tracing method for double-stranded DNA viruses".
<p>The dataset contains mmCIF formatted all-atom models of SU10 and P68 viral genomes and the python script used for refinement of the initial MMB-generated coordinates.</p>
Data from: Quantized crystalline-electromagnetic responses in insulators
Open the record for dataset details and reuse information.
Data from: Anomalous Landau quantization in intrinsic magnetic topological insulators
Open the record for dataset details and reuse information.
Dataset for "The Operator Product Expansion for Radial Lattice Quantization of 3D φ4 Theory"
<h2><strong>Description of the data</strong></h2><p>This dataset is supplement to the paper "The Operator Product Expansion for Radial Lattice Quantization of 3D φ4 Theory". It contains </p><ul><li>the raw data for the partial wave expansion coefficients of the scalar four-point amplitude in the interacting theory for different lattice refinements. The corresponding files are named Rawdata_L=_.json.</li><li>the fit results going into the model averaged lattice results for the OPE coefficients and scaling dimensions of the interacting theory. The corresponding files are named ModelAveraging_L=_.csv.</li><li>the synthetic lattice data for the free theory at different lattice refinements L and cylinder lengths L_t. The corresponding files are named free_cj_L_Lt_.dat.</li></ul><h3><strong>Rawdata_L=_.json</strong></h3><p>The Rawdata_L=_.json files contain data for the partial wave expansion coefficients c_j of the antipodal conformal four-point amplitude that were computed as described in Section IV of the paper. They are given for even j up to j=20 for different values of the lattice refinement L. </p><p>The first two entries in each file are the value of the lattice refinement L as well as the length of the cylinder L_t. </p><p>Thereafter, the partial wave coefficients "c_j" are given for different j as a function of lattice cylinder time t. For each j, we store a list of c_j(t), with the first entry corresponding to t=0, the second to t=1 etc. up until t=L_t/2. Note that the range of t-values is only half of the cylinder length. This is because we used periodic boundary conditions in our simulations and combined the entries for the corresponding times t and L_t-t. </p><p>"c_j_err" then gives the statistical errors of the partial wave expansion coefficients c_j for the different values of j, again as a function of t, starting with t=0 and ending at t=L_t/2+1.</p><p>Lastly, c_conv contains the covariance matrix of the data. Here, we give the covariance matrix of the c_j data for all j, which we have concatenated for this purpose, starting with c_0 at t=0, running to c_0 at t=L_t/2+1 and then continuing with c_2 at t=0 etc until reaching c_20 at t=L_t/2+1. Thus, the entry with index (i, j) will give the covariance of c_{i integer division by L_t/2+1} at time t = i mod L_t/2+1 with c_{j integer division by L_t/2+1} at time t = j mod L_t/2+1. (Especially, the square roots of the diagonal entries give a concatenated list of c_j_err for all j.)</p><h3><strong>ModelAveraging_L_.csv</strong></h3><p>The ModelAveraging_L=_.csv files contain tables of the fit results that go into the final model averaged fit results shown in the paper, i.e. the fits that pass all employed cuts making sure the fits are physical, the parameters are constrained and the fits have an adequate model probability. For a detailed description of the fitting procedure and the employed cuts, see Section IV of the paper. </p><p>Each line of the CSV file should contain 21 entries, separated by commas. Each line corresponds to one fit, and the fits are given in descending order of model probability. For each fit, we give</p><ul><li>"t_0_min": the firstf timeslice we include in this fit for c_0</li><li>"t_2_min": the first timeslice we include in this fit for c_2</li><li>"chi2_dof": the reduced chi^2 value of this fit</li><li>"AIC": the value of the Akaike Information Criterion for this fit</li><li>"prob": the model probability of this fit</li><li>"fe": the fit value for the OPE coefficient f^2_σσε</li><li>"fe_err": the fitting error for the OPE coefficient f^2_σσε</li><li>"De": the fit value for the scaling dimension ∆_ε</li><li>"De_err": the fitting error for the scaling dimension ∆_ε</li><li>"fT": the fit value for the OPE coefficient f^2_σσT</li><li>"fT_err": the fitting error for the OPE coefficient f^2_σσT</li><li>"DT": the fit value for the scaling dimension ∆_T</li><li>"DT_err": the fitting error for the scaling dimension ∆_T</li><li>"fep": the fit value for the OPE coefficient f^2_σσε'</li><li>"fep_err": the fitting error for the OPE coefficient f^2_σσε'</li><li>"Dep": the fit value for the scaling dimension ∆_ε'</li><li>"Dep_err": the fitting error for the scaling dimension ∆_ε'</li><li>"fTp": the fit value for the OPE coefficient f^2_σσT'</li><li>"fTp_err": the fitting error for the OPE coefficient f^2_σσT'</li><li>"DTp": the fit value for the scaling dimension ∆_T'</li><li>"DTp_err": the fitting error for the scaling dimension ∆_T'</li></ul><h3><strong>free_cj_L_Lt_.dat</strong></h3><p>Finally, in the free_cj_L_Lt_.dat files, we give the synthetic lattice data for the free theory calculated on our simplicial lattices by inversion of the quadratic action as described in Section VI of the paper. This data does not have errors, so we simply give the data for the free partial wave expansion coefficients c_j(t) as a function of the lattice time t. For each combination of L and L_t we supply a separate file, with the values of L and L_t written in the filename after the corresponding letter. Each file has 3 columns, the first one indicating the value of j, the second one the time t and the last one the value for c_j(t). The columns are separated by " ". Note that for the free theory, we only calculated c_j(t) for j up to 12.</p>
Charge-4e and Charge-6e Flux Quantization and Higher Charge Superconductivity in Kagome Superconductor Ring Devices
Open the record for dataset details and reuse information.
Non-Majorana-origin of the half-integer conductance quantization elucidated by multi-terminal superconductor-quantum anomalous Hall insulator heterostructure
<p>This folder contains the raw data files used for generating the figures in our preprint (<a href="https://arxiv.org/abs/2411.14903v1">arXiv:2411.14903</a>), "Non-Majorana-origin of the half-integer conductance quantization elucidated by multi-terminal superconductor-quantum anomalous Hall insulator heterostructure", written by A. Uday, G. Lippertz, B. Bhujel, A. A. Taskin, and Y. Ando.</p> <p>The measurement data for each figure are provided in CSV format within the respective folders. Labels and units are specified in the file headers. The figures of the manuscript are also included in PDF format.</p>
Integrals for "Quantum Simulations of Chemistry in First Quantization with any Basis Set"
<p>Integrals used to study scalings and carry out resource estimations. We did not upload the matrices for basis sets with 32000 functions because of large matrix sizes (8 GB per matrix). These integrals correspond to matrix elements of the Hamiltonian in the original basis (Eq. (II.1) and Eq.(III.6)) in this version of the manuscript: (<a href="https://arxiv.org/abs/2408.03145v2">arXiv:2408.03145v2</a> [quant-ph]), and not in the Pauli string representation. </p> <p>UPDATE: in the previous version of matrices in dual plane wave basis, kinetic energy + electron nuclear term gave the correct one-body contribution, but each individual file had an error in it. In this version, we updated the correct kinetic energy and electron nuclear matrices. Notice that electron-nuclear files are just diagonal elements so you need to convert it to a matrix to get correct kinetic + electron nuclear term. Also, dual plane waves electron-electron matrix includes a factor of 1/2 in it. FCIDUMP files are the same as in previous version.</p>
Characterizing and Understanding the Behavior of Quantized Models for Reliable Deployment
<p>quantization benchmark</p>
Observation of plateau regions for zero bias peaks within 5% of the quantized conductance value 2e2/h
<p>This repository contains the raw data and processing Python scripts corresponding to the paper "Observation of plateau regions for zero bias peaks within 5% of the quantized conductance value 2e2/h"</p>
Data for "Half-integer quantized anomalous thermal Hall effect in the Kitaev material candidate alpha-RuCl3"
<p>Data for "Half-integer quantized anomalous thermal Hall effect in the Kitaev material candidate alpha-RuCl3". Several mistakes in the first version are corrected. </p>
Geometry and Integral Files for the manuscript "Quantum Computation for Periodic Solids in Second Quantization"
<p>Geometry and Integral Files for the manuscript "Quantum Computation for Periodic Solids in Second Quantization" <a href="https://doi.org/10.48550/arXiv.2210.02403">https://doi.org/10.48550/arXiv.2210.02403</a></p>
Extended dataset for retracted paper "Quantized Majorana Conductance"
<p>This repository contains an extended dataset of raw data underlying the retracted paper <a href="https://doi.org/10.1038/nature26142">H. Zhang et al, RETRACTED ARTICLE: Quantized Majorana conductance, Nature <strong>556</strong>, 74 (2018)</a></p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.