English

Beyond Chunks and Graphs: Retrieval-Augmented Generation through Triplet-Driven Thinking

Information Retrieval 2025-08-05 v1

Abstract

Retrieval-augmented generation (RAG) is critical for reducing hallucinations and incorporating external knowledge into Large Language Models (LLMs). However, advanced RAG systems face a trade-off between performance and efficiency. Multi-round RAG approaches achieve strong reasoning but incur excessive LLM calls and token costs, while Graph RAG methods suffer from computationally expensive, error-prone graph construction and retrieval redundancy. To address these challenges, we propose T2^2RAG, a novel framework that operates on a simple, graph-free knowledge base of atomic triplets. T2^2RAG leverages an LLM to decompose questions into searchable triplets with placeholders, which it then iteratively resolves by retrieving evidence from the triplet database. Empirical results show that T2^2RAG significantly outperforms state-of-the-art multi-round and Graph RAG methods, achieving an average performance gain of up to 11\% across six datasets while reducing retrieval costs by up to 45\%. Our code is available at https://github.com/rockcor/T2RAG

Keywords

Cite

@article{arxiv.2508.02435,
  title  = {Beyond Chunks and Graphs: Retrieval-Augmented Generation through Triplet-Driven Thinking},
  author = {Shengbo Gong and Xianfeng Tang and Carl Yang and Wei jin},
  journal= {arXiv preprint arXiv:2508.02435},
  year   = {2025}
}

Comments

19 pages

R2 v1 2026-07-01T04:33:22.696Z