{"url":"/dataset/holstep","name":"HolStep","full_name":null,"description_markdown":"HolStep is a dataset based on higher-order logic (HOL) proofs, for the purpose of developing new machine learning-based theorem-proving strategies. \r\n\r\nSource: [HolStep: A Machine Learning Dataset for Higher-order Logic Theorem Proving](/paper/holstep-a-machine-learning-dataset-for-higher)","description_withheld":null,"homepage":"http://cl-informatik.uibk.ac.at/cek/holstep/","introduced_date":null,"introduced_date_note":null,"introduced_by":{"paper":"/paper/holstep-a-machine-learning-dataset-for-higher","title":"HolStep: A Machine Learning Dataset for Higher-order Logic Theorem Proving","first_author":"Cezary Kaliszyk","url":null},"license":{"name":"BSD-3-Clause","url":null},"modalities":[],"tasks":[{"name":"Automated Theorem Proving","url":"/task/automated-theorem-proving","datasets_with_task":"/datasets/task/automated-theorem-proving"},{"name":"Mathematical Proofs","url":"/task/mathematical-proofs","datasets_with_task":"/datasets/task/mathematical-proofs"},{"name":"Dimensionality Reduction","url":"/task/dimensionality-reduction","datasets_with_task":"/datasets/task/dimensionality-reduction"}],"languages":[],"variants":["HolStep (Conditional)","HolStep (Unconditional)","HolStep"],"data_loaders":[],"num_papers_in_archive":10,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/automated-theorem-proving-on-holstep","task":"Automated Theorem Proving","dataset_variant":"HolStep (Conditional)","rows":5,"metrics":["Classification Accuracy"],"first_row_in_archive_order":{"model":"MPNN-DagLSTM","paper":"/paper/improving-graph-neural-network","metrics":{"Classification Accuracy":"0.916"},"code_links":[{"title":"IBM/LogicalFormulaEmbedder","url":"https://github.com/IBM/LogicalFormulaEmbedder"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/automated-theorem-proving-on-holstep-1","task":"Automated Theorem Proving","dataset_variant":"HolStep (Unconditional)","rows":4,"metrics":["Classification Accuracy"],"first_row_in_archive_order":{"model":"FormulaNet","paper":"/paper/premise-selection-for-theorem-proving-by-deep","metrics":{"Classification Accuracy":"0.900"},"code_links":[{"title":"princeton-vl/FormulaNet","url":"https://github.com/princeton-vl/FormulaNet"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/improving-graph-neural-network","title":"Improving Graph Neural Network Representations of Logical Formulae with Subgraph Pooling","date":"2019-11-15","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":0,"samples_unverified":3,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/premise-selection-for-theorem-proving-by-deep","title":"Premise Selection for Theorem Proving by Deep Graph Embedding","date":"2017-09-28","rows_on_this_dataset":4,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/holstep-a-machine-learning-dataset-for-higher","title":"HolStep: A Machine Learning Dataset for Higher-order Logic Theorem Proving","date":"2017-03-01","rows_on_this_dataset":4,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":6,"samples_ran":0,"samples_unverified":6,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":3,"samples_harvested":10,"samples_ran":1,"samples_unverified":9,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":2,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}