datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
qwen36-arithmetic-readouts
Arithmetic Intermediate Readout Sensitivity on Qwen3.6-27B
Arithmetic intermediate detection with a Jacobian lens depends sharply on the prompt token being read and the numeral forms accepted by the scorer. Across 105 order-of-operations items, the recorded hosted-lens responses contain the intermediate at rank 1 on 48 items at the trailing space, versus 4 at the preceding token. On the 25 held-out items with two-digit intermediates, rank-1 detection falls from 13 to 3 when… See the full description on the dataset page: https://huggingface.co/datasets/ec75hash/qwen36-arithmetic-readouts.arithmetic_additionarithmetic_subtractionRWKV-7-ArithmeticRWKV-7-Arithmetic-0.1B 加减法运算模型的训练和测试数据集。
该模型实现基础加减法运算和加减法方程求解功能，能够处理整数部分为 1-12 位、小数部分为 0-6 位的数值，支持中英文数字、全半角格式以及大小写字符的多种表示形式，可实现基础加减法运算和加减法方程求解功能。
训练数据集说明
以下是我们使用的加减法训练数据类型，共包含 30000587 33000147 条单轮加减法 QA 数据，约 1B（1014434168） token。
数据文件名
数据条数
数据说明
示例
ADD_4M
3997733
1. 使用‘全角’、‘中文数字’、‘大写中文数字’随机替换整个数字2. 运算符附近有 1~2 个随机空格3. 含简单自然语言描述/自然语言噪声
{"text": "User: 249476576 减 796580834 还剩多少？\n\nAssistant: -547104258"}
ADD_2M
1999673
1. 使用‘全角’、‘中文数字’、‘大写中文数字’随机替换整个数字2. 运算符附近有 1~2… See the full description on the dataset page: https://huggingface.co/datasets/shoumenchougou/RWKV-7-Arithmetic.mod-arithmetic
Modular Arithmetic Dataset
Synthetic dataset of modular-arithmetic problems of the form a mod b,
paired with the result and a hypothesis about the most suitable tokenizer.
Tokenizer hypothesis
For a mod b where b = 2^k × 5^j (no other prime factors), only the
rightmost max(k, j) digits of a determine the answer, because
10^max(k,j) ≡ 0 (mod b). A tokenizer that groups digits right-to-left
in chunks of that size exposes the relevant information as a single token.
For all… See the full description on the dataset page: https://huggingface.co/datasets/flexitok/mod-arithmetic.TAT-QA-Arithmetic-CoT
Dataset Information
A Chain of Thought (CoT) version of the TAT-QA arithmetic dataset (hosted at https://huggingface.co/datasets/nvidia/ChatQA-Training-Data). The dataset was synthetically generated by prompting Llama3 70B Instruct. The dataset was created as part of our work on Cerebras DocChat - a document-based conversational Q&A model. We observed that initial iterations of our model frequently made errors on arithmetic tasks (such as ConvFinQA) because it was trained on… See the full description on the dataset page: https://huggingface.co/datasets/cerebras/TAT-QA-Arithmetic-CoT.column-arithmetic-ru-synthetic
Column Arithmetic RU Dataset
Синтетический датасет для обучения модели сложению и вычитанию в столбик.
Splits
train.jsonl: основное обучение
eval.jsonl: holdout-оценка
hard.jsonl: трудные случаи с длинными переносами и займами
Hard cases included
9999+1
10000+9999
9090+1010
55555+55555
10999+2
1234+8766
1000-7
10000-9999
50005-49999
8000-1
10101-909
100000-1
99009+991
12000-3456
700000+300001
1002003-998877
Current release status… See the full description on the dataset page: https://huggingface.co/datasets/foxycuter/column-arithmetic-ru-synthetic.adaption-sec-financial-arithmetic-dataset
SEC Financial Arithmetic Dataset — Adaption AutoScientist Challenge
Powered by Adaptive Data — Adaption Labs
What This Dataset Teaches
This dataset trains a model to extract numbers from SEC filing tables and execute verified multi-step arithmetic — every answer is cross-checked against a gold reasoning program:
Task
Source
Example
Table Variable Extraction
FinQA
"From this 10-K table, extract 2021 and 2022 revenue values"
Multi-Step Arithmetic… See the full description on the dataset page: https://huggingface.co/datasets/narendarcodes/adaption-sec-financial-arithmetic-dataset.adaption-arithmetic-algebra-word-problems
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-arithmetic-algebra-word-problems
This dataset features instruction and response pairs containing grade-school arithmetic and algebra word problems paired with step-by-step solutions. Problems cover multi-step arithmetic, percentages, ratios, linear equations, and simple systems solvable in two to five steps. Each completion demonstrates explicit reasoning and concludes with a… See the full description on the dataset page: https://huggingface.co/datasets/vinod-anbalagan/adaption-arithmetic-algebra-word-problems.craft-arithmetic24-spark-budget-study
CraftArithmetic24 and Spark CPU budget study
Twenty-four original synthetic arithmetic questions across six related problem families. Each family contains a base problem, paraphrase, changed-number problem and irrelevant-detail variant. See dataset.jsonl and dataset_metadata.json for prompts, exact labels, version and hash.
The original questions and labels are released under CC0-1.0. The original evaluation code is released under MIT. Model-generated outputs are included as… See the full description on the dataset page: https://huggingface.co/datasets/eruantion87/craft-arithmetic24-spark-budget-study.arithmetic
Арифметические задачи для диалоговой системы
Датасет содержит сэмплы с простыми математическими заданиями примерно такого вида:
- Фонарик Федора работает от 2 батареек, а фонарик Лехи от 6. Сколько батареек нужно фонарикам Федора и Лехи в сумме?
- 2+6=8, столько батареек потребуется.
- Теперь прибавь к результату 469, что получилось?
- 8 плюс 469 равно 477
- Подели на 53, что получилось?
- 9
Основная масса задач связана с арифметическими действиями. Есть некоторое количество задач… See the full description on the dataset page: https://huggingface.co/datasets/inkoziev/arithmetic.basic_arithmeticlemonseed-arithmetic
lemonseed-arithmetic
LemonSeed — multi-digit arithmetic scratchpad (decomposition method).
Contents
arithmetic.jsonl (16000 rows)
Format
JSON Lines (.jsonl), one example per line.
Provenance
Synthetic, generated programmatically for the LemonSeed 1.5B project (by Geramy L. Loveless). Data authored by Michael Anthony Falabella.
Word_Prob_Arithmetic3-digit-arithmetic-scratchpad-traces
Contents
Split
Rows
train
100,000
validation
4,000
test
4,000
total
108,000
Splits are prompt-disjoint — no expression appears in more than one split,
and commutative swaps and trace keys are de-duplicated across splits to prevent
split leakage.
Operation
Rows
×
32,000
÷
32,000
+
22,000
−
22,000
Operands lie in [−999, 999]. Division answers use a fixed DDD.ddd form
(round-half-up to three decimals); division by zero is an atomic <nan>.… See the full description on the dataset page: https://huggingface.co/datasets/vmal/3-digit-arithmetic-scratchpad-traces.Numeric_and_Alpha_Instruct_Arithmeticmath_arithmetic_matrix_rank_test_inmath_arithmetic_matrix_inverse_trainAlpha_Rep_ArithmeticAlpha_Rep_Instructions_Arithmeticmath_arithmetic_matrix_power_level_6arithmetics_mibmath_arithmetic_matrix_determinant_test_inmath_arithmetic_matrix_rank_7B_trainmath_arithmetic_matrix_inverse_level_4math_arithmetic_matrix_power_level_4math_arithmetic_matrix_inverse_level_7Arithmetic
Recursive Arithmetic Transformer training frames
There was no static training file. The model sampled integers online every
step. This dataset replays that sampler with the same Python RNGs used in
train_recursive:
RNG
Seed
What it draws
mix_rng
seed + 91
task mix, operand lengths, integers
offset_rng
seed + 17
Position Coupling origin
seed = 42 for every run. Each train() call resets both RNGs, so later
finetunes are not a continuation of earlier streams.… See the full description on the dataset page: https://huggingface.co/datasets/Amartya77/Arithmetic.math_arithmetic_matrix_power_test_inmath_arithmetic_matrix_svd_test_in