Datasets:
pretty_name: Science QA Samples
license: other
license_name: poseidon-sample-preview
license_link: https://huggingface.co/datasets/psdn-ai/science-qa-samples/blob/main/TERMS.md
tags:
- text
- science
- qa
- education
- samples
configs:
- config_name: default
data_files:
- split: train
path: data/samples.jsonl
task_categories:
- question-answering
size_categories:
- n<1K
Science Q&A Samples
This sample shows structured science question-answer pairs for reviewing subject coverage, difficulty labeling, and answer format before scoping a larger educational dataset.
What This Shows
- Q&A examples across science and math subjects
- Metadata for topic, difficulty, curriculum alignment, and question type
- A view of how text and asset-backed questions are represented
Dataset Specifications
| Field | Value |
|---|---|
| Modality | Text / image Q&A pairs |
| Domain | Science and math education |
| Subjects | Physics, Chemistry, Math, Biology |
| Alignment | JEE, NEET, CET-style curriculum |
| Sample size | 5 rows |
Metadata Coverage
The value is in the structured fields: inputs, outputs, labels, task context, and quality metadata that make the examples useful for evaluation and downstream data planning.
Sample metadata includes: sample identifiers, sample titles, PII status, feature flags, task type, subject, difficulty, curriculum alignment, question text, answer text, correct option, question type, sub-topic, topic.
Sample Scope
The sample includes 5 rows. Larger coverage can be scoped by subject, exam alignment, language, answer style, and media requirements.
For broader coverage or custom requirements, contact hello@psdn.ai or visit psdn.ai/contact.