RecallM: An Adaptable Memory Mechanism with Temporal Understanding for Large Language Models

1 code implementation • 6 Jul 2023 • Brandon Kynoch, Hugo Latapie, Dwane van der Sluis

Unlike previous methods, the RecallM architecture is particularly effective at belief updating and maintaining a temporal understanding of the knowledge provided to it.

Continual Learning In-Context Learning +3