| From Local to Global: A Graph RAG Approach to Query-Focused Summarization |
2024 |
P0 |
pdf-level |
Q1, Q2, Q3, Q4, Q5, Q6, Q8, Q9 |
GraphRAG-Microsoft, # GraphRAG 2025–2026:代表论文、研究脉络与问题演进 |
| HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models |
2024 |
P0 |
pdf-level |
Q1, Q3, Q5, Q6, Q7 |
HippoRAG |
| KG-R1 - Efficient and Transferable Agentic KG-RAG |
2025 |
P0 |
pdf-level |
Q2, Q3, Q4, Q5, Q6 |
KG-R1 - Efficient and Transferable Agentic KG-RAG |
| BubbleRAG - Evidence-Driven RAG for Black-Box KGs |
2026 |
P0 |
pdf-level |
Q1, Q2, Q3, Q5, Q6 |
BubbleRAG - Evidence-Driven RAG for Black-Box KGs |
| CrossAug - Cross-Chunk Graph Augmentation for GraphRAG |
2026 |
P0 |
pdf-level |
Q2, Q3, Q4, Q5, Q6 |
CrossAug - Cross-Chunk Graph Augmentation for GraphRAG |
| DOTRAG - Retrieval-Time Reasoning Along Paths |
2026 |
P0 |
pdf-level |
Q2, Q3, Q4, Q5, Q6 |
DOTRAG - Retrieval-Time Reasoning Along Paths |
| Ex-GraphRAG - Interpretable Evidence Routing for Graph-Augmented LLMs |
2026 |
P0 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
Ex-GraphRAG - Interpretable Evidence Routing for Graph-Augmented LLMs |
| Graph-R1 - Agentic GraphRAG via End-to-end Reinforcement Learning |
2026 |
P0 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
Graph-R1 - Agentic GraphRAG via End-to-end Reinforcement Learning |
| GraphRAG on Consumer Hardware - Local LLMs for Healthcare EHR Schema Retrieval |
2026 |
P0 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
GraphRAG on Consumer Hardware - Local LLMs for Healthcare EHR Schema Retrieval |
| GraphRAG-R1 - Process-Constrained Reinforcement Learning for GraphRAG |
2026 |
P0 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
GraphRAG-R1 - Process-Constrained Reinforcement Learning for GraphRAG |
| HyperGraphPro - Progress-Aware Hypergraph RAG |
2026 |
P0 |
pdf-level |
Q2, Q3, Q4, Q5, Q6 |
HyperGraphPro - Progress-Aware Hypergraph RAG |
| IA-RAG - Interval-Algebra Temporal Reasoning for Dynamic Knowledge Retrieval |
2026 |
P0 |
pdf-level |
Q2, Q3, Q4, Q5, Q6 |
IA-RAG - Interval-Algebra Temporal Reasoning for Dynamic Knowledge Retrieval |
| LLM-Wiki - Retrieval as Reasoning via Self-Evolving Agent-Native Retrieval |
2026 |
P0 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
LLM-Wiki - Retrieval as Reasoning via Self-Evolving Agent-Native Retrieval |
| LiteSemRAG - Lightweight LLM-Free Semantic-Aware Graph Retrieval |
2026 |
P0 |
pdf-level |
Q2, Q3, Q4, Q5, Q6 |
LiteSemRAG - Lightweight LLM-Free Semantic-Aware Graph Retrieval |
| MOSAIC: Query-Aware Exploration Policy Adaptation for GraphRAG |
2026 |
P0 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
MOSAIC |
| PersonalAI 2.0 - Planning Graph Traversal for Personalized LLM Agents |
2026 |
P0 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
PersonalAI 2.0 - Planning Graph Traversal for Personalized LLM Agents |
| Subgraph Reconstruction Attacks on Graph RAG Deployments |
2026 |
P0 |
pdf-level |
Q2, Q3, Q4, Q5, Q6 |
Subgraph Reconstruction Attacks on Graph RAG Deployments |
| Beyond Chunks and Graphs: Retrieval-Augmented Generation through Triplet-Driven Thinking |
— |
P1 |
abstract-level |
Q2, Q3, Q4, Q5, Q8 |
T²RAG |
| ContextRAG - Extraction-Free Hierarchical Graph Construction |
— |
P1 |
repair-needed |
Q2, Q5 |
ContextRAG - Extraction-Free Hierarchical Graph Construction |
| Dynamic Chunking and Selection for Reading Comprehension of Ultra-Long Context in Large Language Models |
— |
P1 |
abstract-level |
Q2, Q3, Q5, Q8, Q9 |
Dynamic-Chunking-for-Ultra-Long-Context |
| GeAR: Graph-enhanced Agent for Retrieval-augmented Generation |
— |
P1 |
abstract-level |
Q3, Q4, Q5, Q8, Q9 |
GeAR |
| Generative Dense Retrieval: Memory Can Be a Burden |
— |
P1 |
abstract-level |
Q1, Q3, Q5, Q6, Q8 |
Generative-Dense-Retrieval-Memory-Burden |
| Graph-based RAG 研究脉络阅读报告 |
— |
P1 |
abstract-level |
Q2, Q3, Q5, Q7, Q8 |
GraphRAG研究脉络阅读报告, GraphRAG研究脉络阅读报告 |
| Injecting Structured Biomedical Knowledge into Language Models - Continual Pretraining vs GraphRAG |
— |
P1 |
repair-needed |
Q10, Q5 |
Injecting Structured Biomedical Knowledge into Language Models - Continual Pretraining vs GraphRAG |
| M3KG-RAG - Multi-hop Multimodal Knowledge Graph-enhanced RAG |
— |
P1 |
repair-needed |
Q10, Q5 |
M3KG-RAG - Multi-hop Multimodal Knowledge Graph-enhanced RAG |
| Query-Aware-GNN |
— |
P1 |
repair-needed |
Q10, Q4, Q5 |
Query-Aware-GNN |
| ReinPool: Reinforcement Learning Pooling Multi-Vector Embeddings for Retrieval System |
— |
P1 |
abstract-level |
Q2, Q3, Q5, Q6, Q8 |
RL-Pooling-Multi-Vector-Embeddings |
| Rethinking Chunk Size for Long-Document Retrieval: A Multi-Dataset Analysis |
— |
P1 |
abstract-level |
Q2, Q3, Q5, Q8, Q9 |
Rethinking-Chunk-Size-for-Long-Document-Retrieval |
| SUBQRAG - Sub-Question Driven Dynamic Graph RAG |
— |
P1 |
repair-needed |
Q5 |
SUBQRAG - Sub-Question Driven Dynamic Graph RAG |
| Why Neighborhoods Matter - Traversal Context and Provenance in Agentic GraphRAG |
— |
P1 |
repair-needed |
Q4, Q5, Q6 |
Why Neighborhoods Matter - Traversal Context and Provenance in Agentic GraphRAG |
| 研究报告:从节点相关性到可验证证据集合 |
— |
P1 |
abstract-level |
Q2, Q3, Q4, Q5, Q6, Q7 |
研究报告:从节点相关性到可验证证据集合, 检索技术启发性研究阅读报告, 检索技术启发性研究阅读报告 |
| Think-on-Graph: Deep and Responsible Reasoning of Large Language Model on Knowledge Graph |
2024 |
P1 |
pdf-level |
Q2, Q3, Q5, Q6, Q7 |
Think-on-Graph-ToG |
| E2GraphRAG: Streamlining Graph-based RAG for High Efficiency and Effectiveness |
2025 |
P1 |
pdf-level |
Q1, Q2, Q3, Q5, Q7 |
E²GraphRAG |
| HopRAG - Multi-Hop Reasoning for Logic-Aware Retrieval-Augmented Generation |
2025 |
P1 |
pdf-level |
Q2, Q3, Q5, Q8, Q9 |
HopRAG |
| Inference Scaled GraphRAG: Improving Multi Hop Question Answering on Knowledge Graphs |
2025 |
P1 |
pdf-level |
Q3, Q4, Q5, Q6, Q7 |
Inference-Scaled-GraphRAG |
| LINEAR RAG: Linear Graph Retrieval Augmented Generation on Large-Scale Corpora |
2025 |
P1 |
pdf-level |
Q1, Q2, Q3, Q5, Q7 |
LinearRAG |
| A2RAG - Adaptive Agentic Graph Retrieval |
2026 |
P1 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
A2RAG - Adaptive Agentic Graph Retrieval |
| CS-RAG - Robust GraphRAG over Imperfect KGs |
2026 |
P1 |
pdf-level |
Q2, Q3, Q4, Q5, Q6, Q7, Q8, Q9 |
CS-RAG - Robust GraphRAG over Imperfect KGs, C2RAG |
| CacheRAG - Semantic Caching for KGQA |
2026 |
P1 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
CacheRAG - Semantic Caching for KGQA |
| Do We Still Need GraphRAG - Agentic Search Benchmark |
2026 |
P1 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
Do We Still Need GraphRAG - Agentic Search Benchmark |
| FastInsight - Fast and Insightful Retrieval via Fusion Operators for Graph RAG |
2026 |
P1 |
pdf-level |
Q2, Q3, Q4, Q5, Q8 |
FastInsight - Fast and Insightful Retrieval via Fusion Operators for Graph RAG |
| GraphRAG-Router - Cost-Efficient Routing over GraphRAGs and LLMs |
2026 |
P1 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
GraphRAG-Router - Cost-Efficient Routing over GraphRAGs and LLMs |
| NaviRAG - Active Knowledge Navigation for RAG |
2026 |
P1 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
NaviRAG - Active Knowledge Navigation for RAG |
| PRoH - Dynamic Planning over Knowledge Hypergraphs |
2026 |
P1 |
pdf-level |
Q2, Q3, Q4, Q5, Q6 |
PRoH - Dynamic Planning over Knowledge Hypergraphs |
| Reasoning Bottleneck in Graph-RAG - Structured Prompting and Context Compression |
2026 |
P1 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
Reasoning Bottleneck in Graph-RAG - Structured Prompting and Context Compression |
| STAR - Semantic-Tuned and Tail-Adaptive Retriever for GraphRAG |
2026 |
P1 |
pdf-level |
Q1, Q2, Q3, Q4, Q5 |
STAR - Semantic-Tuned and Tail-Adaptive Retriever for GraphRAG |
| TCAR-Gen - Temporal Graph Retrieval with Evidence Fusion |
2026 |
P1 |
pdf-level |
Q1, Q2, Q3, Q5, Q6 |
TCAR-Gen - Temporal Graph Retrieval with Evidence Fusion |