Awesome-Code-LLM is a curated research collection that catalogs language modeling studies focused on code and software engineering tasks, along with associated datasets.
The collection addresses the need for a comprehensive overview of the rapidly expanding field of code-focused language models. It organizes research across multiple dimensions: foundational models and pretraining approaches, adaptation of existing language models to code, instruction fine-tuning and reinforcement learning techniques, the intersection of coding with reasoning tasks, and specialized applications for low-resource or domain-specific languages. The repository structures papers into categories including code search, code generation, code understanding, and code agents, making it possible to navigate the landscape of code LLM research systematically.
Developers and researchers building code-related AI systems should use this collection to understand the state of the field and identify relevant prior work. It suits anyone working on code generation, code search, code understanding, or reasoning systems that leverage code. The repository is particularly valuable for those exploring how existing language models can be adapted for code tasks, or for understanding specialized applications like code agents and interactive coding systems. The collection includes papers from major venues and conferences, providing access to both established and emerging research directions.
The project maintains active curation with regular additions of papers from major conferences and venues. Contributors are invited to report missing papers, miscategorizations, or incomplete reference information through the issue tracker. The repository includes supplementary resources beyond papers, such as links to code implementations and model releases associated with featured research.
