Papers
arxiv:2602.01965

Breaking the Static Graph: Context-Aware Traversal for Robust Retrieval-Augmented Generation

Published on Feb 2
· Submitted by
Kwun Hang
on Feb 6
Authors:
,
,
,
,
,

Abstract

CatRAG addresses limitations in retrieval-augmented generation by introducing a query-adaptive framework that improves multi-hop reasoning through symbolic anchoring, dynamic edge weighting, and key-fact passage enhancement.

AI-generated summary

Recent advances in Retrieval-Augmented Generation (RAG) have shifted from simple vector similarity to structure-aware approaches like HippoRAG, which leverage Knowledge Graphs (KGs) and Personalized PageRank (PPR) to capture multi-hop dependencies. However, these methods suffer from a "Static Graph Fallacy": they rely on fixed transition probabilities determined during indexing. This rigidity ignores the query-dependent nature of edge relevance, causing semantic drift where random walks are diverted into high-degree "hub" nodes before reaching critical downstream evidence. Consequently, models often achieve high partial recall but fail to retrieve the complete evidence chain required for multi-hop queries. To address this, we propose CatRAG, Context-Aware Traversal for robust RAG, a framework that builds on the HippoRAG 2 architecture and transforms the static KG into a query-adaptive navigation structure. We introduce a multi-faceted framework to steer the random walk: (1) Symbolic Anchoring, which injects weak entity constraints to regularize the random walk; (2) Query-Aware Dynamic Edge Weighting, which dynamically modulates graph structure, to prune irrelevant paths while amplifying those aligned with the query's intent; and (3) Key-Fact Passage Weight Enhancement, a cost-efficient bias that structurally anchors the random walk to likely evidence. Experiments across four multi-hop benchmarks demonstrate that CatRAG consistently outperforms state of the art baselines. Our analysis reveals that while standard Recall metrics show modest gains, CatRAG achieves substantial improvements in reasoning completeness, the capacity to recover the entire evidence path without gaps. These results reveal that our approach effectively bridges the gap between retrieving partial context and enabling fully grounded reasoning. Resources are available at https://github.com/kwunhang/CatRAG.

Community

Paper author Paper submitter

This paper addresses a fundamental limitation in graph-based retrieval-augmented generation (RAG) systems, which we characterize as the "Static Graph Fallacy." While recent methods have successfully utilized Knowledge Graphs (KGs) to capture multi-hop dependencies, the reliance on fixed transition probabilities often results in semantic drift, where retrieval is diverted toward high-degree "hub" nodes rather than relevant evidence.

CatRAG introduces a context-aware traversal framework that transforms the static KG into a query-adaptive navigation structure. By integrating symbolic anchoring and dynamic edge weighting, the system effectively prunes irrelevant paths and amplifies those aligned with the query’s specific intent. A key finding of our work is that while standard recall metrics show modest gains, there is a significant improvement in "reasoning completeness"—the ability to recover the entire evidence chain without gaps. This shift from partial context retrieval to grounded reasoning paths is a necessary step for robust multi-hop RAG.

We look forward to discussing the implications of dynamic graph steering and how these techniques might scale to increasingly large and heterogeneous knowledge structures.

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2602.01965 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2602.01965 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2602.01965 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.