The field of natural language processing is witnessing significant developments in retrieval-augmented generation (RAG) and graph-based reasoning. Recent research has focused on enhancing large language models (LLMs) with external knowledge sources, such as graphs and databases, to improve their performance on knowledge-intensive tasks. One of the key directions is the integration of graph structures into RAG frameworks, enabling more precise and interpretable reasoning. Another important area of research is the development of novel architectures and mechanisms for efficient and adaptive retrieval, such as iterative retrieval and multi-agent systems. These advancements have the potential to revolutionize various applications, including question answering, text generation, and decision support systems. Noteworthy papers in this area include PIR-RAG, which introduces a practical system for privacy-preserving RAG, and G-reasoner, a unified framework that integrates graph and language foundation models for reasoning over diverse graph-structured knowledge. Additionally, papers like MIXRAG and Think-on-Graph 3.0 have proposed innovative approaches to graph-based RAG, demonstrating state-of-the-art performance on various benchmarks.
Advancements in Retrieval-Augmented Generation and Graph-Based Reasoning
Sources
PIR-RAG: A System for Private Information Retrieval in Retrieval-Augmented Generation
HetaRAG: Hybrid Deep Retrieval-Augmented Generation across Heterogeneous Data Stores
MIXRAG : Mixture-of-Experts Retrieval-Augmented Generation for Textual Graph Understanding and Question Answering
Think-on-Graph 3.0: Efficient and Adaptive LLM Reasoning on Heterogeneous Graphs via Multi-Agent Dual-Evolving Context Retrieval
GraphSearch: An Agentic Deep Searching Workflow for Graph Retrieval-Augmented Generation
AEGIS: Authentic Edge Growth In Sparsity for Link Prediction in Edge-Sparse Bipartite Knowledge Graphs
Beyond Textual Context: Structural Graph Encoding with Adaptive Space Alignment to alleviate the hallucination of LLMs
Rethinking and Benchmarking Large Language Models for Graph Reasoning
G-reasoner: Foundation Models for Unified Reasoning over Graph-structured Knowledge
PhysicsMinions: Winning Gold Medals in the Latest Physics Olympiads with a Coevolutionary Multimodal Multi-Agent System
ActorDB: A Unified Database Model Integrating Single-Writer Actors, Incremental View Maintenance, and Zero-Trust Messaging
Beyond Static Retrieval: Opportunities and Pitfalls of Iterative Retrieval in GraphRAG
ReTAG: Retrieval-Enhanced, Topic-Augmented Graph-Based Global Sensemaking
Type-Less yet Type-Aware Inductive Link Prediction with Pretrained Language Models
OntoAligner Meets Knowledge Graph Embedding Aligners
Private Information Retrieval over Graphs
Exploring Network-Knowledge Graph Duality: A Case Study in Agentic Supply Chain Risk Analysis
TAG-EQA: Text-And-Graph for Event Question Answering via Structured Prompting Strategies
REBot: From RAG to CatRAG with Semantic Enrichment and Graph Routing
Are LLMs Better GNN Helpers? Rethinking Robust Graph Learning under Deficiencies with Iterative Refinement