Core Concepts
Current large language models (LLMs) lack a robust episodic memory system, struggling to recall the sequential order of information, especially when relying on parametric or retrieval-augmented memory mechanisms.
Pink1, M., Vo2, V. A., Wu1, Q., Mu3, J., Turek2, J., Hasson4,5, U., Norman4,5, K. A., Michelmann6, S., Huth3, A., & Toneva1, M. (2024). Assessing Episodic Memory in LLMs with Sequence Order Recall Tasks. arXiv. https://arxiv.org/abs/2410.08133v1
This research paper investigates the capacity of large language models (LLMs) to exhibit episodic memory, focusing on their ability to recall the sequential order of information.