Inference Scaled GraphRAG: Improving Multi Hop Question Answering on Knowledge Graphs
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915358139482112 |
|---|---|
| author | Thompson, Travis Lim, Seung-Hwan Liu, Paul He, Ruoying Xu, Dongkuan |
| author_facet | Thompson, Travis Lim, Seung-Hwan Liu, Paul He, Ruoying Xu, Dongkuan |
| contents | Large Language Models (LLMs) have achieved impressive capabilities in language understanding and generation, yet they continue to underperform on knowledge-intensive reasoning tasks due to limited access to structured context and multi-hop information. Retrieval-Augmented Generation (RAG) partially mitigates this by grounding generation in retrieved context, but conventional RAG and GraphRAG methods often fail to capture relational structure across nodes in knowledge graphs. We introduce Inference-Scaled GraphRAG, a novel framework that enhances LLM-based graph reasoning by applying inference-time compute scaling. Our method combines sequential scaling with deep chain-of-thought graph traversal, and parallel scaling with majority voting over sampled trajectories within an interleaved reasoning-execution loop. Experiments on the GRBench benchmark demonstrate that our approach significantly improves multi-hop question answering performance, achieving substantial gains over both traditional GraphRAG and prior graph traversal baselines. These findings suggest that inference-time scaling is a practical and architecture-agnostic solution for structured knowledge reasoning with LLMs |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_19967 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Inference Scaled GraphRAG: Improving Multi Hop Question Answering on Knowledge Graphs Thompson, Travis Lim, Seung-Hwan Liu, Paul He, Ruoying Xu, Dongkuan Computation and Language Artificial Intelligence Large Language Models (LLMs) have achieved impressive capabilities in language understanding and generation, yet they continue to underperform on knowledge-intensive reasoning tasks due to limited access to structured context and multi-hop information. Retrieval-Augmented Generation (RAG) partially mitigates this by grounding generation in retrieved context, but conventional RAG and GraphRAG methods often fail to capture relational structure across nodes in knowledge graphs. We introduce Inference-Scaled GraphRAG, a novel framework that enhances LLM-based graph reasoning by applying inference-time compute scaling. Our method combines sequential scaling with deep chain-of-thought graph traversal, and parallel scaling with majority voting over sampled trajectories within an interleaved reasoning-execution loop. Experiments on the GRBench benchmark demonstrate that our approach significantly improves multi-hop question answering performance, achieving substantial gains over both traditional GraphRAG and prior graph traversal baselines. These findings suggest that inference-time scaling is a practical and architecture-agnostic solution for structured knowledge reasoning with LLMs |
| title | Inference Scaled GraphRAG: Improving Multi Hop Question Answering on Knowledge Graphs |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2506.19967 |