---
title: "AriGraph: Learning Knowledge Graph World Models with Episodic Memory for LLM Agents"
authors:
  - "Petr Anokhin"
  - "Nikita Semenov"
  - "Artyom Sorokin"
  - "Dmitry Evseev"
  - "Andrey Kravchenko"
  - "Mikhail Burtsev"
  - "Evgeny Burnaev"
doi: "10.24963/ijcai.2025/2"
url: "https://ijcai.org/proceedings/2025/2"
pdf: "https://ijcai.org/proceedings/2025/0002.pdf"
venue: "Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence"
year: 2025
pages: "12-20"
track: "Main Track"
keywords:
  - "Natural Language Processing: NLP: Information retrieval and text mining"
  - "Agent-based and Multi-agent Systems: MAS: Multi-agent learning"
  - "Agent-based and Multi-agent Systems: MAS: Multi-agent planning"
  - "Data Mining: DM: Knowledge graphs and knowledge base completion"
---

# AriGraph: Learning Knowledge Graph World Models with Episodic Memory for LLM Agents

Petr Anokhin, Nikita Semenov, Artyom Sorokin, Dmitry Evseev, Andrey Kravchenko, Mikhail Burtsev, Evgeny Burnaev

*Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence*, pages 12-20. DOI: [10.24963/ijcai.2025/2](https://doi.org/10.24963/ijcai.2025/2)

## Abstract

Advancements in the capabilities of Large Language Models (LLMs) have created a promising foundation for developing autonomous agents. With the right tools, these agents could learn to solve tasks in new environments by accumulating and updating their knowledge. Current LLM-based agents process past experiences using a full history of observations, summarization, retrieval augmentation. However, these unstructured memory representations do not facilitate the reasoning and planning essential for complex decision-making. In our study, we introduce AriGraph, a novel method wherein the agent constructs and updates a memory graph that integrates semantic and episodic memories while exploring the environment. We demonstrate that our Ariadne LLM agent, consisting of the proposed memory architecture augmented with planning and decision-making, effectively handles complex tasks within interactive text game environments difficult even for human players. Results show that our approach markedly outperforms other established memory methods and strong RL baselines in a range of problems of varying complexity. Additionally, AriGraph demonstrates competitive performance compared to dedicated knowledge graph-based methods in static multi-hop question-answering.