Datasets:
metadata
annotations_creators: []
language:
- en
language_creators: []
license: []
multilinguality:
- monolingual
pretty_name: adaption-math_misconception_match
size_categories:
- 10K<n<100K
source_datasets:
- original
tags:
- adaption
- instruction-tuning
- academic-education
- math
task_categories: []
task_ids: []
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-math_misconception_match
This dataset contains pairs of math problems, student wrong answers, and candidate misconceptions. The task is to determine if the specific misconception logically explains the student's error. Each sample provides the question context, the incorrect response, and a labeled classification of 'MATCH' or 'NO_MATCH'.
Dataset size
There are 19,996 data points in this dataset. This is an instruction tuning dataset.
Quality of Remastered Dataset
The final quality is B, with a relative quality improvement of 21.4%.
Domain
- Academic-education (96%)
- Math (4%)
Language
- English (100%)
Tone
- Analytical (100%)
Evaluation Results
Quality Gains:
Grade Improvement:
Percentile Chart:
