Datasets:
license: mit
task_categories:
- text-generation
language:
- en
tags:
- math
- reasoning
- mathematics
- Cot
- ToT
- TrainOfThought
pretty_name: Train Of Thought - Mathematics Version 1.0.0
size_categories:
- 10K<n<100K
configs:
- config_name: default
data_files:
- split: train
path: ToT-Math-26k.json
Train of Thought: Mathematics V1
Synthetic English question–answer pairs with written reasoning for mathematical problem solving, from elementary topics to advanced mathematics.
Contents
Format: English text; JSON records with question, answer, id, and metadata (difficulty, topic, and generated reasoning).
Topics: Arithmetic, geometry, algebra, real analysis, abstract algebra, mathematical physics, and applications in fields such as data science and finance.
Creation and limitations
The Genesis II pipeline generated the dataset using Qwen, Phi, and Llama models with structured prompting and format checks. The written reasoning was designed to mimic o1-style explanations.
Synthetic proofs, calculations, and explanations may be incorrect or incomplete. Independently verify results before using them for evaluation, instruction, or consequential work.
Use
from datasets import load_dataset
dataset = load_dataset("mattwesney/ToT-Math-V1")
Authenticate with an account that has access to this private dataset.
License: MIT. Version: 1.0.0.
Citation
Matthew Wesney (2025), Train of Thought: Mathematics V1. Hugging Face dataset.