PaperSense

AI-Powered Research Paper Assistant for Chrome

PaperSense is an in-context research assistant Chrome Extension designed to help researchers quickly decide whether a paper is worth reading, understand key methods, and identify transferable ideas — all powered by on-device AI or Gemini API.

✨ Features

📄 Paper Snapshot

Get a quick overview of any research paper:

• One-liner summary of the paper's contribution
• Relevance rating based on your research interests
• Suggested reading strategy (skim, deep read, or skip)

💡 Usage Insights

Understand how to apply the paper's ideas:

• Abstracts the core contribution
• Suggests 2-3 concrete ways to apply it to your research domain

🔍 Highlight → Explain

Select any text in a paper, right-click, and get a contextual explanation:

• Explain: Technical explanation of selected text
• Translate: Translate selected text to your preferred language

📸 Visual Analysis (Screenshot)

Capture and analyze figures, diagrams, or equations:

• Right-click anywhere → "Explain Visual Area with PaperSense"
• Draw a selection box around the visual content
• Get AI-powered explanation of the visual

🌐 Multi-language Support

• English
• 繁體中文 (Traditional Chinese)
• 简体中文 (Simplified Chinese)
• 日本語 (Japanese)

🚀 Installation

1. Clone this repository or download the source code
2. Open Chrome and navigate to chrome://extensions
3. Enable Developer mode (toggle in the top right)
4. Click Load unpacked and select the paper-sense folder
5. Click the PaperSense icon or open the Chrome Side Panel to start!

⚙️ AI Configuration

PaperSense supports two AI backends:

Option 1: Gemini Nano (On-device AI) — Recommended

Enjoy private, free, and offline inference directly in your browser.

Requirements:

• Chrome 127 or newer
• macOS 13+ / Windows 10+ / Linux
• At least 22GB free disk space
• GPU with >4GB VRAM, or CPU with 16GB+ RAM

Setup Steps:

1. Open chrome://flags in your browser

2. Enable the following flags:

   Flag	Setting
   #optimization-guide-on-device-model	Enabled BypassPerfRequirement
   #prompt-api-for-gemini-nano	Enabled
   #prompt-api-for-gemini-nano-multimodal-input	Enabled (for screenshot feature)

3. Click Relaunch to restart Chrome

4. Open PaperSense Settings and select Gemini Nano

5. Wait for the model to download (the extension will guide you)

Option 2: Gemini API

For more powerful inference or if your hardware doesn't support Nano.

1. Get an API Key from Google AI Studio
2. Open PaperSense Settings
3. Select Gemini API and paste your key
4. (Optional) Change the model name (default: gemini-2.5-flash-lite)
5. Click Save Settings

📖 How to Use

Basic Workflow

1. Configure Interests: Open Settings and describe your research area
2. Visit a Paper: Navigate to arXiv, IEEE, ACL, or other academic sites
3. Analyze: Click "Analyze This Paper" for Snapshot and Usage Insights
4. Deep Dive: Use the chat feature to ask follow-up questions

Context Menu Features

Right-click on any webpage to access:

Menu Item	Description
Explain with PaperSense	Explain selected text
Translate with PaperSense	Translate selected text
Explain Visual Area	Capture and analyze a screenshot region

Feature Comparison

Feature	Gemini Nano	Gemini API
Paper Analysis	✓	✓
Text Explanation	✓	✓
Text Translation	✓	✓
Visual/Screenshot Analysis	✓*	✓
PDF Analysis	✗	✓
Offline Support	✓	✗
Privacy (On-device)	✓	✗

*Requires enabling #prompt-api-for-gemini-nano-multimodal-input flag

🛠️ Development

Project Structure

paper-sense/
├── manifest.json          # Extension manifest
├── src/
│   ├── background.js      # Service worker
│   ├── content.js         # Content script for page interaction
│   ├── services/          # AI service modules
│   │   ├── ai.js          # Main AI interface
│   │   ├── gemini-api.js  # Gemini API implementation
│   │   └── gemini-nano.js # Gemini Nano implementation
│   ├── sidepanel/         # Side panel UI
│   ├── options/           # Options page
│   ├── prompt/            # Prompt templates
│   ├── styles/            # CSS stylesheets
│   └── icons/             # Extension icons
└── README.md

Tech Stack

• Chrome Extension Manifest V3
• Chrome Built-in AI (Prompt API) for on-device inference
• Gemini API for cloud inference
• KaTeX for LaTeX rendering
• Vanilla JavaScript (no framework dependencies)

📜 License

This project is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).

You are free to:

• Share — copy and redistribute the material in any medium or format
• Adapt — remix, transform, and build upon the material

Under the following terms:

• Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made.
• NonCommercial — You may not use the material for commercial purposes.

For more details, see: CC BY-NC 4.0 License

🙏 Acknowledgments

• Google Chrome Built-in AI for the Prompt API
• Google AI Studio for Gemini API access
• KaTeX for beautiful math rendering

Built with ❤️ for researchers