{"success":true,"data":{"author":"Gregor Kasieczka, Tilman Plehn, Jennifer Thompson, Michael Russel (original); Parquet conversion by dl4phys","bucketAccessType":"public","bucketName":"dl4phys/top_tagging","bucketPrefix":"","category":"Physics","cloudProvider":"huggingface","createdAt":"2026-11-14T14:15:23.588Z","description":"The Top Quark Tagging Reference Dataset of Kasieczka, Plehn and colleagues, the community benchmark behind the 2019 comparison paper 'The Machine Learning Landscape of Top Taggers' (SciPost Physics). Monte Carlo events from 14 TeV proton-proton collisions were generated with Pythia8 and the Delphes ATLAS detector card, without multiple interactions or pile-up, clustered into anti-kT R=0.8 jets with pT 550 to 650 GeV; signal jets are hadronic tops and background is QCD. Each row has the four-momenta (E, px, py, pz) of the leading 200 jet constituents sorted by transverse momentum and zero-padded, the truth top four-momentum, and the label is_signal_new. This Parquet conversion, made for the Hugging Face 'Deep Learning for Particle Physicists' course, keeps the original splits: 1,211,000 train, 403,000 validation and 404,000 test jets; files are train.parquet, validation.parquet and test.parquet, about 2.3 GB in total (measured 2026-11-14). Load with datasets, pandas or pyarrow. The original Zenodo release (10.5281/zenodo.2603256) is CC-BY-4.0; cite Kasieczka et al. 2019.","displayTitle":"Top Quark Tagging Reference Dataset (Parquet)","doi":"10.5281/zenodo.2603256","id":"f217e791-b237-5005-be73-76a930fcbac0","license":"CC-BY-4.0","licenseEvidenceUrl":"https://zenodo.org/records/2603256","objectCountEstimate":7,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/1902.09914","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"dl4phys/top_tagging\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"dl4phys/top_tagging\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=dl4phys, license:cc-by-4.0, size_categories:1M<n<10M, format:parquet, downloads=221"],"format":"mixed","generatedAt":"2026-11-14T14:15:29.281Z","layout":{"example":"test.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"test.parquet","truncatedAt":512}],"sampleObjects":["create-datasets.ipynb","test.parquet","train.parquet","validation.parquet"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\n---\n# Dataset Card for Top Quark Tagging\n\n## Table of Contents\n- [Dataset Card for Top Quark Tagging](#dataset-card-for-top-quark-tagging)\n  - [Table of Contents](#table-of-contents)\n  - [Dataset Description](#dataset-description)\n    - [Dataset Summary](#dataset-summary)\n    - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)\n  - [Dataset Structure](#dataset-structure)\n    - [Data Instances](#data-instances)\n    - [Data Fields](#data-fields)\n    - [Data Splits](#data-splits)\n    - [Licensing Information](#licensing-information)\n    - [Citation Information](#citation-information)\n    - [Contributions](#contributions)\n\n## Dataset Description\n\n- **Homepage:** https://zenodo.org/record/2603256\n- **Paper:** https://arxiv.org/abs/1902.09914\n- **Point of Contact:** [Gregor Kasieczka](gregor.kasieczka@uni-hamburg.de)\n\n### Dataset Summary\n\nTop Quark Tagging is a dataset of Monte Carlo simulated events produced by proton-proton collisions at the Large Hadron Collider. The top-quark signal and mixed quark-gluon background jets are produced with Pythia8 with its default tune for a center-of-mass energy of 14 TeV. Multiple interactions and pile-up are ignored. The leading 200 jet constituent four-momenta \\\\( (E, p_x, p_y, p_z) \\\\) are stored, with zero-padding applied to jets with fewer than 200 constituents.\n\n### Supported Tasks and Leaderboards\n\n- `tabular-classification`: The dataset can be used to train a model for tabular binary classification, which consists in predicting whether an event is produced from a top signal or quark-gluon background. Success on this task is typically measured by achieving a *high* [accuracy](https://huggingface.co/metrics/accuracy) and AUC score.\n\n## Dataset Structure\n\n### Data Instances\n\nEach instance in the dataset consists of the four-momenta of the leading 200 jet constituents, sorted by \\\\(p_T\\\\). For jets with fewer than 200 constituents, zero-padding is applied. The four-momenta of the top-quark are also provided, along with a label in the `is_signal_new` column to indicate whether the event stems from a top-quark (1) or QCD background (0). An example instance looks as follows:\n\n```\n{'E_0': 474.0711364746094,\n 'PX_0': -250.34703063964844,\n 'PY_0': -223.65196228027344,\n 'PZ_0': -334.73809814453125,\n  ...\n 'E_199': 0.0,\n 'PX_199': 0.0,\n 'PY_199': 0.0,\n 'PZ_199': 0.0,\n 'truthE': 0.0,\n 'truthPX': 0.0,\n 'truthPY': 0.0,\n 'truthPZ': 0.0,\n 'ttv': 0,\n 'is_signal_new': 0}\n```\n\n### Data Fields\n\nThe fields in the dataset have the following meaning:\n\n- `E_i`: the energy of jet constituent \\\\(i\\\\).\n- `PX_i`: the \\\\(x\\\\) component of the jet constituent's momentum\n- `PY_i`: the \\\\(y\\\\) component of the jet constituent's momentum\n- `PZ_i`: the \\\\(z\\\\) component of the jet constituent's momentum\n- `truthE`: the energy of the top-quark\n- `truthPX`: the \\\\(x\\\\) component of the top quark's momentum\n- `truthPY`: the \\\\(y\\\\) component of the top quark's momentum\n- `truthPZ`: the \\\\(z\\\\) component of the top quark's momentum\n- `ttv`: a flag that indicates which split (train, validation, or test) that a jet belongs to. Redundant since each split is provided as a separate dataset\n- `is_signal_new`: the label for each jet. A 1 indicates a top-quark, while a 0 indicates QCD background.\n\n### Data Splits\n\n|                  |   train | validation |   test |\n|------------------|--------:|-----------:|-------:|\n| Number of events | 1211000 |     403000 | 404000 |\n\n### Licensing Information\n\nThis dataset is released under the [Creative Commons Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/legalcode) license.\n### Citation Information\n\n```\n@dataset{kasieczka_gregor_2019_2603256,\n  author       = {Kasieczka, Gregor and\n                  Plehn, Tilman and\n                  Thompson, Jennifer and\n                  Russel, Michael},\n  title        = {Top Quark Tagging Reference Dataset},\n  month        = mar,\n  year         = 2019,\n  publisher    = {Zenodo},\n  version      = {v0 (2018\\_03\\_27)},\n  doi          = {10.5281/zenodo.2603256},\n  url          = {ht","encoding":"utf8","key":"README.md","name":"README.md","size":4232,"truncated":true}],"sourceObjectKey":"test.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"dl4phys/top_tagging\")\nprint(ds)","starterSamplePath":"test.parquet","totalObjects":7},"summary":"The Top Quark Tagging Reference Dataset of Kasieczka, Plehn and colleagues, the community benchmark behind the 2019 comparison paper 'The Machine Learning Landscape of Top…","title":"Top Quark Tagging Reference Dataset (Parquet)","updatedAt":"2026-11-14T17:26:33.420Z","visibility":"public","websiteUrl":"https://zenodo.org/records/2603256"}}