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.

翻译：摘要：大语言模型在引入新知识时面临遗忘或高成本重新训练的困境。本文提出DYNA轻量级框架，通过时序知识图谱增强冻结的大语言模型，其中事件作为节点，时序关系作为带时间戳的有向边。该图谱作为可更新的外部记忆模块，在查询时通过随机游走与中心性度量检索相关节点，进而增强大语言模型输出。在三个时序回忆任务中，DYNA相比微调将灾难性遗忘降低约7%，相比标准RAG将时序排序准确率提升约5%。图谱聚类系数与检索性能呈正相关，证实图谱结构的重要性。贡献包括：(1) 将情景记忆建模为时序知识图谱；(2) 无需重新训练的模型增强方案；(3) 以图谱属性预测检索性能。