HAGE: Harnessing Agentic Memory via RL-Driven Weighted Graph Evolution

TL;DR

HAGE introduces a dynamic, query-conditioned graph traversal approach for agentic memory in LLMs, improving reasoning accuracy and efficiency by jointly optimizing memory routing and representations with reinforcement learning.

Contribution

It presents a novel weighted multi-relational memory framework with RL-based training that enhances retrieval adaptability and reasoning in large language models.

Findings

Improved long-horizon reasoning accuracy over state-of-the-art methods.

Achieved a better accuracy-efficiency trade-off in memory retrieval.

Demonstrated effectiveness through empirical evaluations.

Abstract

Memory retrieval in agentic large language model (LLM) systems is often treated as a static lookup problem, relying on flat vector search or fixed binary relational graphs. However, fixed graph structures cannot capture the varying strength, confidence, and query-dependent relevance of relationships between events. In this paper, we propose HAGE, a weighted multi-relational memory framework that reconceptualizes retrieval as sequential, query-conditioned traversal over a unified relational memory graph. Memory is organized as relation-specific graph views over shared memory nodes, where each edge is associated with a trainable relation feature vector encoding multiple relational signals. Given a query, an LLM-based classifier identifies the relational intent, and a routing network dynamically modulates the corresponding dimensions of the edge embedding. Traversal scores are computed via a learned combination of semantic similarity and these query-conditioned edge representations. This allows memory traversal to prioritize high-utility relational paths while softly suppressing noisy or weakly relevant connections. Beyond adaptive traversal, HAGE further introduces a reinforcement learning-based training framework that jointly optimizes routing behavior and edge representations using downstream tasks. Finally, empirical results demonstrate improved long-horizon reasoning accuracy and a favorable accuracy-efficiency trade-off compared to state-of-the-art agentic memory systems. Our code is available at https://github.com/FredJiang0324/HAGE_MVPReview.