This repository serves as a curated, daily-updated engineering guide and resource hub for in-context learning (ICL) and prompt engineering. It targets researchers, developers, and practitioners aiming to master Large Language Models (LLMs) like ChatGPT, GPT-3, and FlanT5, providing a structured approach to leveraging these technologies for advanced applications.

How It Works

The project organizes a vast collection of resources into distinct categories: Papers (surveys, ICL, prompt engineering techniques, agents, multimodal prompts, foundation models), a Playground for LLM experimentation, specific Prompt Engineering guides, ChatGPT prompt examples, and an LLM Usage guide featuring LangChain tutorials. This structured approach facilitates systematic learning and application of cutting-edge LLM techniques.

Quick Start & Requirements

• Installation: No explicit installation instructions are provided, suggesting the repository is primarily a curated collection of links and information rather than a runnable software package.
• Prerequisites: Access to LLMs (e.g., ChatGPT, GPT-3, FlanT5) and potentially LangChain for the usage guide.
• Resources: Links to papers (arXiv, DOI), GitHub repositories, and tutorials are provided throughout the README.

Highlighted Details

• Comprehensive categorization of papers covering ICL, prompt engineering, agents, RAG, evaluation, multimodal prompts, and foundation models.
• Includes practical guides for prompt engineering techniques and ChatGPT prompt examples.
• Offers a LangChain tutorial for LLM usage and deployment.
• Regularly updated with recent research and news in the LLM space.

Maintenance & Community

• Maintained by EgoAlpha Lab.
• Contact available via helloegoalpha@gmail.com for discussions and collaborations.

Licensing & Compatibility

• The repository itself does not specify a license. Individual linked resources (papers, code) will have their own licenses.

Limitations & Caveats

This repository is a curated list of resources and does not provide executable code or a unified framework. Users must independently access and integrate the linked papers, playgrounds, and tools. The "Playground" section lacks specific implementation details or direct links.