---
task_categories:
- question-answering
language:
- en
tags:
- circuit
- physics
- circuit_benchmark
- hierarchical_benchmark
- circuit_understanding
size_categories:
- 1K<n<10K
license: mit
arxiv: 2509.22339
---
# CircuitSense

This dataset is a comprehensive multimodal circuit question-answering benchmark designed to evaluate visual reasoning and problem-solving capabilities across three main domains: Perception, Analysis, and Design. The dataset contains structured question-answer pairs with accompanying visual content, targeting different engineering cognitive levels and reasoning tasks.


### Dataset Description

* Homepage: https://circuitsense-benchmark.github.io
* Repository: https://github.com/armanakbari/CircuitSense/tree/main
* Paper: https://arxiv.org/pdf/2509.22339
* Point of Contact: akbari.ar@northeastern.edu



## Dataset Structure
The dataset is organized into three primary folders, each containing different types of visual reasoning tasks:


```
├── Perception/
│   └── [level_folders]/
│       └── q#/
│           ├── q#_question.txt
│           ├── q#_image.png
│           ├── q#_ta.txt
│           ├── q#-a.txt
│           └── q#_mc.txt
├── Analysis/
│   ├── curated/
│   │   └── [level_folders]/
│   │       └── q#/
│   │           ├── q#_question.txt
│   │           ├── q#_image.png
│   │           ├── q#_ta.txt
│   │           ├── q#-a.txt
│   │           └── q#_mc.txt
│   └── synthetic/
│       └── [level_folders]/
│           └── q#/
│               ├── q#_question.txt
│               ├── q#_image.png
│               ├── q#_ta.txt
│               ├── q#-a.txt
│               └── q#_mc.txt
└── Design/
    └── [level_folders]/
        └── q#/
            ├── q#_question.txt
            ├── q#_image.png
            ├── q#_ta.txt
            ├── q#-a.txt
            └── q#_mc.txt
```


### File Descriptions

Each question folder contains the following standardized files:

* `q#_question.txt`: Contains the question text/prompt
* `q#_image.png`: Visual content associated with the question
* `q#_ta.txt`: Teaching assistant notes or detailed explanations
* `q#-a.txt`: Ground truth answer(s)
* `q#_mc.txt`: Multiple choice options (if applicable)