- P
MiniF2F Dataset
- paperswithcode.com
- opendatalab.com
Updated Aug 14, 2024
Share
Facebook
Twitter
EmailClick to copy linkLink copiedCiteKunhao Zheng; Jesse Michael Han; Stanislas Polu (2024). MiniF2F Dataset [Dataset]. https://paperswithcode.com/dataset/minif2fDataset updatedAug 14, 2024AuthorsKunhao Zheng; Jesse Michael Han; Stanislas PoluDescriptionMiniF2F is a dataset of formal Olympiad-level mathematics problems statements intended to provide a unified cross-system benchmark for neural theorem proving. The miniF2F benchmark currently targets Metamath, Lean, and Isabelle and consists of 488 problem statements drawn from the AIME, AMC, and the International Mathematical Olympiad (IMO), as well as material from high-school and undergraduate mathematics courses.
- h
minif2f_test
- huggingface.co
Updated Apr 28, 2025ShareFacebookTwitterEmailClick to copy linkLink copiedCiteProject-Numina (2025). minif2f_test [Dataset]. https://huggingface.co/datasets/AI-MO/minif2f_testDataset updatedApr 28, 2025Dataset authored and provided byProject-NuminaLicenseApache License, v2.0https://www.apache.org/licenses/LICENSE-2.0
License information was derived automaticallyDescriptionMiniF2F
Dataset UsageThe evaluation results of Kimina-Prover presented in our work are all based on this MiniF2F test set.
ImprovementsWe corrected several erroneous formalizations, since the original formal statements could not be proven. We list them in the following table. All our improvements are made based on the MiniF2F test set provided by DeepseekProverV1.5, which applies certain modifications to the original dataset to adapt it to the Lean 4.… See the full description on the dataset page: https://huggingface.co/datasets/AI-MO/minif2f_test.
- h
minif2f_solving
- huggingface.co
Updated May 8, 2025ShareFacebookTwitterEmailClick to copy linkLink copiedCiteQi Liu (2025). minif2f_solving [Dataset]. https://huggingface.co/datasets/purewhite42/minif2f_solvingDataset updatedMay 8, 2025AuthorsQi LiuLicenseApache License, v2.0https://www.apache.org/licenses/LICENSE-2.0
License information was derived automaticallyDescriptionDataset Card for MiniF2F-Solving
This benchmark is part of the official implementation of Beyond Theorem Proving: Formulation, Framework and Benchmark for Formal Problem-Solving. Our research focuses on:
What is problem-solving? Beyond proving known targets, how can process-verified problem-solving be conducted inside existing formal theorem proving (FTP) environments?
ContributionA principled formulation of problem-solving as a deterministic Markov decision process;… See the full description on the dataset page: https://huggingface.co/datasets/purewhite42/minif2f_solving.
miniF2F-Graded
- zenodo.org
bin, json, pngUpdated Jan 31, 2025ShareFacebookTwitterEmailClick to copy linkLink copiedCiteAnonymous; Anonymous (2025). miniF2F-Graded [Dataset]. http://doi.org/10.5281/zenodo.14776138png, json, binAvailable download formatsUnique identifierhttps://doi.org/10.5281/zenodo.14776138Dataset updatedJan 31, 2025AuthorsAnonymous; AnonymousLicenseAttribution 4.0 (CC BY 4.0)https://creativecommons.org/licenses/by/4.0/
License information was derived automaticallyDescriptionMiniF2F-Graded(./miniF2F-Graded.json) builds upon miniF2F by introducing additional metrics for each theorem: Difficulty, Discrimination, and Difficulty Grading. These metrics are calculated based on the actual performance of LLMs in proving the theorems, making them a more accurate reflection of difficulty from the perspective of LLMs.Please refer to ./README.md for more information.- h
DSP1.5RL-minif2f-sampling_4
- huggingface.co
Updated Nov 10, 2024ShareFacebookTwitterEmailClick to copy linkLink copiedCiteAutores (2024). DSP1.5RL-minif2f-sampling_4 [Dataset]. https://huggingface.co/datasets/autores/DSP1.5RL-minif2f-sampling_4CroissantCroissant is a format for machine-learning datasets. Learn more about this at mlcommons.org/croissant.Dataset updatedNov 10, 2024Dataset authored and provided byAutoresDescriptionautores/DSP1.5RL-minif2f-sampling_4 dataset hosted on Hugging Face and contributed by the HF Datasets community
- h
dataset_0
- huggingface.co
Updated Apr 13, 2024+ more versionsShareFacebookTwitterEmailClick to copy linkLink copiedCiteXXX (2024). dataset_0 [Dataset]. https://huggingface.co/datasets/xyy888/dataset_0Dataset updatedApr 13, 2024AuthorsXXXDescriptionMiniF2F is a formal mathematics benchmark (translated across multiple formal systems) consisting of exercise statements from olympiads (AMC, AIME, IMO) as well as high-school and undergraduate maths classes. This dataset contains formal statements in Isabelle. Each statement is paired with an informal statement and an informal proof, as described in Draft, Sketch, Prove [Jiang et al 2023]. The problems in this dataset use the most recent facebookresearch/miniF2F commit on July 3, 2023.
Not seeing a result you expected?
Learn how you can add new datasets to our index.
FacebookTwitterEmailMiniF2F is a dataset of formal Olympiad-level mathematics problems statements intended to provide a unified cross-system benchmark for neural theorem proving. The miniF2F benchmark currently targets Metamath, Lean, and Isabelle and consists of 488 problem statements drawn from the AIME, AMC, and the International Mathematical Olympiad (IMO), as well as material from high-school and undergraduate mathematics courses.