Autonomous Knowledge Graph Exploration with Adaptive Breadth-Depth Retrieval

2026-01-20

AI Summary

This work addresses the trade-off between breadth coverage and depth reasoning in multi-hop knowledge graph retrieval, where existing approaches are either limited to shallow traversal or rely on fragile seed entities. The authors propose ARK (Adaptive Retrieval over Knowledge graphs), the first training-free, agent-driven retrieval framework that enables large language models to dynamically adapt to query types—linguistic or relational—through a dual-tool mechanism combining global lexical search and single-hop neighborhood exploration. This design allows autonomous control over retrieval breadth and depth, eliminating dependence on predefined hop counts or seed nodes. Leveraging label-agnostic imitation learning, the tool-use trajectories of a large teacher model are distilled into an 8B-parameter student model. On the STaRK benchmark, ARK achieves 59.1% Hit@1 and 67.4% MRR, outperforming baselines by up to 31.4%; the distilled model improves Hit@1 by 7.0, 26.6, and 13.5 percentage points across three datasets while retaining up to 98.5% of the teacher’s performance.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Search and Optimization: Learning to Search
• Knowledge Representation and Reasoning: Computational Complexity of Reasoning

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Retrieving evidence for language model queries from knowledge graphs requires balancing broad search across the graph with multi-hop traversal to follow relational links. Similarity-based retrievers provide coverage but remain shallow, whereas traversal-based methods rely on selecting seed nodes to start exploration, which can fail when queries span multiple entities and relations. We introduce ARK: Adaptive Retriever of Knowledge, an agentic KG retriever that gives a language model control over this breadth-depth tradeoff using a two-operation toolset: global lexical search over node descriptors and one-hop neighborhood exploration that composes into multi-hop traversal. ARK alternates between breadth-oriented discovery and depth-oriented expansion without depending on a fragile seed selection, a pre-set hop depth, or requiring retrieval training. ARK adapts tool use to queries, using global search for language-heavy queries and neighborhood exploration for relation-heavy queries. On STaRK, ARK reaches 59.1% average Hit@1 and 67.4 average MRR, improving average Hit@1 by up to 31.4% and average MRR by up to 28.0% over retrieval-based and agentic training-free methods. Finally, we distill ARK's tool-use trajectories from a large teacher into an 8B model via label-free imitation, improving Hit@1 by +7.0, +26.6, and +13.5 absolute points over the base 8B model on AMAZON, MAG, and PRIME datasets, respectively, while retaining up to 98.5% of the teacher's Hit@1 rate.

Problem

• Knowledge Graph Retrieval
• Breadth-Depth Tradeoff
• Multi-hop Reasoning
• Query-Evidence Retrieval
• Language Model Queries

Innovation

• adaptive retrieval
• knowledge graph exploration
• breadth-depth tradeoff
• agentic retrieval
• imitation distillation