2026-07-08 08:57:50 +00:00

114 lines
3.3 KiB
Markdown

# Minerva-Math
## Overview
Minerva-Math is a benchmark designed to evaluate advanced mathematical and quantitative reasoning capabilities of language models. It consists of 272 challenging problems sourced primarily from MIT OpenCourseWare courses, covering university and graduate-level STEM subjects.
## Task Description
- **Task Type**: Advanced STEM Problem Solving
- **Input**: University/graduate-level mathematical or scientific problem
- **Output**: Step-by-step solution with final answer
- **Difficulty**: University to graduate level
## Key Features
- 272 challenging problems from MIT OpenCourseWare
- Covers advanced subjects: solid-state chemistry, astronomy, differential equations, special relativity
- University and graduate-level difficulty
- Tests deep mathematical and scientific reasoning
- Problems require multi-step quantitative reasoning
## Evaluation Notes
- Default configuration uses **0-shot** evaluation
- Answers should be formatted within `\boxed{}` for proper extraction
- Uses LLM-as-judge for complex answer evaluation
- Problems may require domain-specific knowledge (physics, chemistry, etc.)
- Designed to test the upper limits of model reasoning capabilities
## Properties
| Property | Value |
|----------|-------|
| **Benchmark Name** | `minerva_math` |
| **Dataset ID** | [knoveleng/Minerva-Math](https://modelscope.cn/datasets/knoveleng/Minerva-Math/summary) |
| **Paper** | N/A |
| **Tags** | `Math`, `Reasoning` |
| **Metrics** | `acc` |
| **Default Shots** | 0-shot |
| **Evaluation Split** | `train` |
## Data Statistics
| Metric | Value |
|--------|-------|
| Total Samples | 272 |
| Prompt Length (Mean) | 492.27 chars |
| Prompt Length (Min/Max) | 129 / 1069 chars |
## Sample Example
**Subset**: `default`
```json
{
"input": [
{
"id": "715a0fa1",
"content": "Each of the two Magellan telescopes has a diameter of $6.5 \\mathrm{~m}$. In one configuration the effective focal length is $72 \\mathrm{~m}$. Find the diameter of the image of a planet (in $\\mathrm{cm}$ ) at this focus if the angular diameter of the planet at the time of the observation is $45^{\\prime \\prime}$.\nPlease reason step by step, and put your final answer within \\boxed{}."
}
],
"target": "Start with:\n\\[\ns=\\alpha f \\text {, }\n\\]\nwhere $s$ is the diameter of the image, $f$ the focal length, and $\\alpha$ the angular diameter of the planet. For the values given in the problem:\n\\[\ns=\\frac{45}{3600} \\frac{\\pi}{180} 7200=\\boxed{1.6} \\mathrm{~cm}\n\\]",
"id": 0,
"group_id": 0,
"metadata": {
"type": "Introduction to Astronomy (8.282J Spring 2006)",
"idx": 0
}
}
```
## Prompt Template
**Prompt Template:**
```text
{question}
Please reason step by step, and put your final answer within \boxed{{}}.
```
## Usage
### Using CLI
```bash
evalscope eval \
--model YOUR_MODEL \
--api-url OPENAI_API_COMPAT_URL \
--api-key EMPTY_TOKEN \
--datasets minerva_math \
--limit 10 # Remove this line for formal evaluation
```
### Using Python
```python
from evalscope import run_task
from evalscope.config import TaskConfig
task_cfg = TaskConfig(
model='YOUR_MODEL',
api_url='OPENAI_API_COMPAT_URL',
api_key='EMPTY_TOKEN',
datasets=['minerva_math'],
limit=10, # Remove this line for formal evaluation
)
run_task(task_cfg=task_cfg)
```