A Machine With Human-Like Memory Systems – NewsEverything | Technology

Cognitive science states that the human memory system is composed of semantic and episodic memory systems. Semantic memory has to do with general world knowledge, while episodic has to do with one’s personal memory.

Computer memory module. Image credit: Lenharth Systems via Stocksnap, CC0 Public Domain

Inspired by this observation, a recent paper published on arXiv.org models an agent that has both semantic and episodic memory systems.

Researchers designed and released a challenging environment where an agent has to properly learn how to encode, store and retrieve memories to maximize rewards. It is shown that an agent with both memory systems answers the questions more successfully than that using only one of the two memory systems.

It is also shown that when an agent is pretrained with commonsense knowledge, it outperforms the one that is not pretrained. Moreover, it is demonstrated that when an agent collaborates with another agent or human, it leads to better performance.

Inspired by the cognitive science theory, we explicitly model an agent with both semantic and episodic memory systems, and show that it is better than having just one of the two memory systems. In order to show this, we have designed and released our own challenging environment, “the Room”, compatible with OpenAI Gym, where an agent has to properly learn how to encode, store, and retrieve memories to maximize its rewards. The Room environment allows for a hybrid intelligence setup where machines and humans can collaborate. We show that two agents collaborating with each other results in better performance than one agent acting alone. We have open-sourced our code and models at this https URL.

Research paper: Kim, T., Cochez, M., Francois-Lavet, V., Neerincx, M., and Vossen, P., “A Machine With Human-Like Memory Systems”, 2022. Link: https://arxiv.org/abs/2204.01611

You can offer your link to a page which is relevant to the topic of this post.