# DYNA : Dynamic Episodic Memory Networks for Augmenting Large Language Models with Temporal Knowledge Graphs in Continuous Learning

arXiv: https://arxiv.org/abs/2606.15778 · HTML: https://mlsanity.com/paper/2606.15778/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 35 references that are in the map.
- First posted on arXiv 14 Jun 2026.

## Builds on (most cited first, 18 of 18)

- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Retrieval-Augmented Generation for Large Language Models: A Survey](https://mlsanity.com/paper/2312.10997/) (2023)
- [Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection](https://mlsanity.com/paper/2310.11511/) (2023)
- [An Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning](https://mlsanity.com/paper/2308.08747/) (2023)
- [Open Domain Question Answering Using Early Fusion of Knowledge Bases and Text](https://mlsanity.com/paper/1809.00782/) (2018)
- [Deep Bidirectional Language-Knowledge Graph Pretraining](https://mlsanity.com/paper/2210.09338/) (2022)
- [Know-Evolve: Deep Temporal Reasoning for Dynamic Knowledge Graphs](https://mlsanity.com/paper/1705.05742/) (2017)
- [Continual Learning for Large Language Models: A Survey](https://mlsanity.com/paper/2402.01364/) (2024)
- [AriGraph: Learning Knowledge Graph World Models with Episodic Memory for LLM Agents](https://mlsanity.com/paper/2407.04363/) (2024)
- [Temporal Knowledge Graph Reasoning with Historical Contrastive Learning](https://mlsanity.com/paper/2211.10904/) (2022)
- [Memory3: Language Modeling with Explicit Memory](https://mlsanity.com/paper/2407.01178/) (2024)
- [A Survey on Temporal Knowledge Graph: Representation Learning and Applications](https://mlsanity.com/paper/2403.04782/) (2024)
- [Inductive Logical Query Answering in Knowledge Graphs](https://mlsanity.com/paper/2210.08008/) (2022)
- [Semantic HELM: A Human-Readable Memory for Reinforcement Learning](https://mlsanity.com/paper/2306.09312/) (2023)
- [ConPET: Continual Parameter-Efficient Tuning for Large Language Models](https://mlsanity.com/paper/2309.14763/) (2023)
- [Neurocoder: Learning General-Purpose Computation Using Stored Neural Programs](https://mlsanity.com/paper/2009.11443/) (2020)
- [Can LLMs Alleviate Catastrophic Forgetting in Graph Continual Learning? A Systematic Study](https://mlsanity.com/paper/2505.18697/) (2025)

## Abstract (arXiv)

Large Language Models (LLMs) struggle to incorporate new knowledge without forgetting or costly retraining. We propose DYNA, a lightweight framework that augments a frozen LLM with a temporal knowledge graph where events are nodes and temporal relations are directed, timestamped edges. The graph serves as an external, updatable memory. At query time, DYNA retrieves relevant nodes via random walks and centrality measures, then augments the LLM's response. Evaluated on three temporal recall tasks, DYNA reduces catastrophic forgetting by ~7% compared to fine-tuning and improves temporal ordering by ~5% over standard RAG. Higher graph clustering coefficients correlate with better retrieval, showing that graph structure matters. Contributions: (1) episodic memory as temporal KG, (2) retraining-free LLM augmentation, (3) graph properties as predictors of retrieval performance.