University Entrance Exam Physics Multi-Step Reasoning Q&A Dataset
#multi-step reasoning #quantitative information extraction #step-by-step problem solving #university entrance exam preparation #physics problem solving #physics education
- 5000
- 0.4GB
- JSON
- CATL
MOBIUSI INC
Updated:2026-10-19
AI Analysis & Value Prop
This dataset provides question-and-answer text for university entrance exam physics problems, including the original question, key physical quantities and constraints, ordered reasoning steps, and the final answer. It focuses on solving problems through multiple steps, from extracting information in the prompt to connecting physical laws and completing calculations or judgments. Use it for reasoning fine-tuning of physics QA models, solution-skills evaluation, and intelligent education applications.
Dataset Insights
Sample Examples
{}
Technical Specifications
| Field | Type | Description |
|---|---|---|
| file_name | string | File name |
| final_answer | string | The final answer, including units or a concise explanation when appropriate. |
| question_text | string | The original exam question, including any stated conditions, values, and constraints. |
| reasoning_steps | array | Ordered solution steps that connect relevant physical laws and calculations to the result. |
| extracted_quantities | array | Physical quantities, values, units, and constraints identified from the question. Represent each entry as a concise text string. |
Compliance Statement
| Authorization Type | Proprietary - Commercial AI Training License (No Redistribution) |
| Commercial Use | Requires exclusive subscription or authorization contract (monthly or per-invocation charging) |
| Privacy and Anonymization | No PII, no real company names, simulated scenarios follow industry standards |
| Compliance System | Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs |
Frequently Asked Questions
- What does the dataset contain?
- Each record is organized around a physics problem and includes the original question, extracted physical quantities and constraints, ordered reasoning steps, and a final answer. This structure supports training and analysis from condition identification through solution.
- Which capabilities can it support?
- It is suitable for multi-step physics reasoning, multi-hop question answering, solution-step generation, and question answering based on physical laws. It can support reasoning SFT and capability evaluation for related models.
- How can I verify the topic coverage and annotation quality?
- The dataset focuses on university entrance exam physics problems and reasoning-oriented solutions. Confirm the item count, topic coverage, answer validation process, and specific quality metrics from the available samples and dataset documentation.
- How can I obtain the data, and what are the licensing terms?
- Please refer to the product page or provider's delivery documentation for access methods, available formats, and licensing terms. Before use, verify whether the license covers your intended purposes, such as training, evaluation, or commercial applications.
Can't find the data you need?
Post a request and let data providers reach out to you.
Get this Dataset
Verified for Enterprise UseCite this Work
@dataset{Mobiusi2026,
title={University Entrance Exam Physics Multi-Step Reasoning Q&A Dataset},
author={MOBIUSI INC},
year={2026},
url={https://www.mobiusi.com/datasets/d1d53c5db84cbf66ebf93fd62a2858f2?dataset_scene_id=7},
urldate={2026-10-19},
keywords={university entrance exam physics dataset,physics multi-step reasoning,physics question answering data,physics problem solving dataset,multi-hop QA dataset,physics reasoning training,reasoning SFT data,educational AI dataset},
version={1.0}
}Using this in research? Please cite us.
placeholder
placeholder
placeholder
placeholder
placeholder
placeholder
placeholder