Back to Arena

PathRAG

by BUPT-GAMMA

System Card

OrganizationBUPT-GAMMA
Released2024-01
Architecturegraph-rag / Path-based relational retrieval
DetailsHybrid of semantic vector, graph traversal, and entity-centric retrieval over entity-edge graphs. Full-stack (React frontend with graph viz, FastAPI backend, SQLite).
Parameters—
Domainrag-retrievalknowledge-graph
Open SourceYes
PaperView Paper
CodeRepository
path-basedgraph-traversalvisualizationhybrid

Capability Profile

Benchmark Scores

6 of 14 benchmarks
Long-Context Retrieval
2/5
RULER
73.670p
NIAH
no data
LooGLE
no data
LongBench
603p
∞Bench
no data
Multi-Turn Recall
0/2
LoCoMo
no data
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
73.837p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
7579p
HotpotQA
7044p
Agent Task Memory
0/1
AgentBench-Mem
no data
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
1/1
RAGAS
71.667p
Sources:PathRAG paper (arXiv:2502.14902); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)PathRAG paper (arXiv:2502.14902); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)PathRAG paper (arXiv:2502.14902); evaluated on RAGAS: Automated Evaluation of Retrieval-Augmented Generation (Exploding Gradients, 2309)PathRAG paper (arXiv:2502.14902); evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)PathRAG paper (arXiv:2502.14902); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)PathRAG paper (arXiv:2502.14902); evaluated on RULER: What's the Real Context Size of Your Long-Context Language Models (NVIDIA, 2404)