https://github.com/para-droid-ai/scratchpad
Scratchpad
https://github.com/para-droid-ai/scratchpad
logic model reasoning scratchpad
Last synced: 5 months ago
JSON representation
Scratchpad
- Host: GitHub
- URL: https://github.com/para-droid-ai/scratchpad
- Owner: para-droid-ai
- License: mit
- Created: 2024-05-23T14:57:46.000Z (11 months ago)
- Default Branch: main
- Last Pushed: 2024-11-15T23:07:28.000Z (5 months ago)
- Last Synced: 2024-11-16T00:18:13.924Z (5 months ago)
- Topics: logic, model, reasoning, scratchpad
- Homepage:
- Size: 42 MB
- Stars: 59
- Watchers: 3
- Forks: 5
- Open Issues: 6
-
Metadata Files:
- Readme: README.md
- License: license.txt
Awesome Lists containing this project
- jimsghstars - para-droid-ai/scratchpad - Scratchpad (Others)
README
# Scratchpad
Scratchpad is a user-defined framework designed for a sophisticated AI assistant capable of deep reasoning and providing insightful responses to user inquiries. The framework works as a form of "user/machine" calibration, with a focus on user-context and user-intent. This README provides an overview, installation instructions, usage guidelines, and contribution information.
## Table of Contents
- [Overview](#overview)
- [Features](#features)
- [Showcase](#showcase)
- [Installation](#installation)
- [Usage](#usage)
- [Contributing](#contributing)
- [License](#license)
## Overview
Scratchpad aims to facilitate advanced reasoning by systematically processing user inputs through various cognitive operations. It dissects the intent behind user queries and formulates logical pathways to reach accurate conclusions.
## Features
- **Focus on User Intent and Context**: Breaks down user intent and context to provide more relevant and accurate responses.
- **Dynamic Contextual Adaptation**: Adapts to changes in user context dynamically, ensuring ongoing relevance in responses.
- **Transparent Reasoning "Blackbox"**: Establishes a "blackbox" for the model's reasoning, enabling users to verify the model's logic and confirm its understanding of user context.
- **Multi-Step Reasoning**: Identifies key elements, analyzes user perspectives, and applies cognitive operations like abstraction, comparison, and inference.
- **Calibration of User Intent**: Aims to calibrate the model to align closely with the user's intent and context, enhancing the relevance and accuracy of responses.
- **Advanced Logic**: Utilizes advanced reasoning to connect the dots and provide insightful responses.
- **Error Detection and Correction**: Detects and corrects potential errors in reasoning, enhancing response reliability.
- **User Feedback Integration**: Integrates user feedback into the reasoning process, improving performance over time.
- **Interactive Learning**: Refines reasoning capabilities through continuous user interaction.
- **Scalable Reasoning Processes**: Maintains performance and accuracy across complex queries and diverse topics.
- **Clarity and Accuracy**: Strives for clear and accurate reasoning, demonstrating advanced artificial intelligence capabilities.
## Showcase
### Links
- [Scratchpad Canvas Best Practices](https://www.perplexity.ai/page/chain-of-thought-reasoning-via-22CYSxmhTMSFr1gJIXM4dg)
- [Deep Dive into the Scratchpad Framework](https://www.perplexity.ai/page/scratchpad-ai-reasoning-framew-790vL5qORlyvX7VSwMYmzg)
- [Reddit Link with Demo GIFs](https://www.reddit.com/r/perplexity_ai/comments/1fm55ha/using_cot_canvas_via_the_complexity_browser/)
- [Podcast Update!](https://open.spotify.com/episode/2CLWkrUKJ1sBf8Li8tigQ7?si=_Z2_MOcaQWmya2T9kw9oBg)
### Demos
- **Image Analysis:**
*Caption: A breakdown of four sky images showcasing the analysis capabilities.*
- **Scratchpad in Action:**
*Caption: Demonstration of using CPLX with Scratchpad Canvas from, from the PPLX homepage.*
- **Creating the Scratchpad Collection in PPLX:**
*Caption: Step-by-step guide to creating a Scratchpad collection for PPLX.*
- **Model Adaptation:**
*Caption: Demonstration of the model's adaptation to the Scratchpad framework.*
## Installation
- For AI Profile, copy/paste the prompt from: [AI Profile Prompt](https://github.com/para-droid-ai/scratchpad/blob/main/prompt_AI-Profile.md)
- For PPLX Spaces, copy/paste the prompt from: [PPLX Space Prompt](https://github.com/para-droid-ai/scratchpad/blob/main/prompt_space.md)
## Usage
Once Scratchpad is added to your AI Profile/PPLX Space (via Space Prompt or File), it will enhance your AI's reasoning capabilities. Here’s how it works:
1. **Automatic Activation**:
- Scratchpad will be automatically triggered during interactions, enhancing the AI's ability to reason and provide insightful responses.
2. **Behind the Scenes**:
- The framework will process user inputs through cognitive operations like abstraction, comparison, and inference without manual intervention.
3. **Output Enhancement**:
- Responses will be structured and logical, aiming for clarity and accuracy in reasoning.
4. **Cognitive Operations**:
- Automatically identifies key elements, restates questions, analyzes perspectives, and applies advanced reasoning processes to generate high-quality answers.
By integrating Scratchpad into your AI Profile/PPLX Space, the framework seamlessly improves the AI's performance, ensuring sophisticated and insightful responses.
## Contributing
We welcome contributions to enhance Scratchpad. To contribute:
1. **Fork the repository.**
2. **Create a new branch:**
```bash
git checkout -b feature-branch
```
3. **Make your changes and commit them:**
```bash
git commit -m "Description of changes"
```
4. **Push to the branch:**
```bash
git push origin feature-branch
```
5. **Open a pull request.**
For questions or further guidance, feel free to contact the maintainers.
## License
Scratchpad is licensed under the MIT License. See the [LICENSE](https://github.com/para-droid-ai/scratchpad/blob/main/LICENSE) file for more details.
---
This README.md provides a clear overview, installation steps, usage instructions, and contribution guidelines, making it easier for users to understand and utilize the Scratchpad framework.