HANRAG (Heuristic Accurate Noise-resistant Retrieval-Augmented Generation) is an open-source implementation of the framework from the paper arXiv:2509.09713v1. It tackles robust RAG when queries are noisy or require multi-hop reasoning.
Focus
- Noise-resistant retrieval: Heuristics and filtering so retrieval stays useful under imperfect or ambiguous queries.
- Multi-hop handling: Design that supports chaining retrieval and generation steps for complex questions.
- LangChain / LangGraph: Built so it fits into existing RAG pipelines and can be extended with custom graphs.
The code is on GitHub and can be used as a reference implementation or integrated into production RAG stacks.