Skip to yearly menu bar Skip to main content


BayesPCN: A Continually Learnable Predictive Coding Associative Memory

Jinsoo Yoo · Frank Wood

Hall J (level 1) #424

Keywords: [ bayesian inference ] [ associative memory ] [ machine learning ] [ continual learning ] [ predictive coding ]

Abstract: Associative memory plays an important role in human intelligence and its mechanisms have been linked to attention in machine learning. While the machine learning community's interest in associative memories has recently been rekindled, most work has focused on memory recall ($read$) over memory learning ($write$). In this paper, we present BayesPCN, a hierarchical associative memory capable of performing continual one-shot memory writes without meta-learning. Moreover, BayesPCN is able to gradually forget past observations ($forget$) to free its memory. Experiments show that BayesPCN can recall corrupted i.i.d. high-dimensional data observed hundreds to a thousand ``timesteps'' ago without a large drop in recall ability compared to the state-of-the-art offline-learned parametric memory models.

Chat is not available.