This repository serves as a curated collection of papers and resources related to Large Language Models (LLMs), focusing on foundational research and evaluation methodologies. It is intended for researchers, engineers, and practitioners in the NLP and AI fields seeking a comprehensive overview of LLM advancements, particularly concerning models like OpenAI's GPT series and Meta's Llama.

How It Works

The repository organizes a vast number of academic papers, categorized by LLM capabilities and research areas such as reasoning, instruction tuning, retrieval augmentation, and multimodal applications. It leverages an automated script powered by Auto-Bibfile to maintain and regenerate its README from BibTeX entries and JSON data, ensuring a structured and up-to-date compilation of relevant literature.

Quick Start & Requirements

• Installation: No direct installation is required as this is a resource repository.
• Prerequisites: Access to the internet to view the linked papers and resources.
• Resources: The README itself acts as the primary resource, linking to numerous papers and research topics.

Highlighted Details

• Extensive categorization of LLM research, covering over 50 distinct sub-fields.
• Includes papers on foundational LLM techniques, evaluation metrics, and emerging applications.
• Features a dedicated section on "In-Context Learning" with numerous sub-topics and related papers.
• Organizes research by specific LLM capabilities like reasoning, generation, and adaptation.

Maintenance & Community

The repository is maintained by Runzhe Wang and Shenyu Zhang, with contributions from a dedicated team of paper collectors and organizers, including Guilin Qi and Xiaofang Qi.

Licensing & Compatibility

The repository itself does not specify a license, but it links to external academic papers, each with its own licensing and usage terms.

Limitations & Caveats

This repository is a collection of links to external research papers and does not provide any executable code or models. The sheer volume of papers means it is a starting point for exploration rather than an exhaustive, annotated guide.