POEM

By bringing together researchers from psychology, cognitive neuroscience, and computational modeling, the goal of POEM is to better understand predictive processing with context-free (semantic) or context-linked (episodic) memory as internal models. The relation is considered in a bidirectional way, that is, the generation of prediction is based on memory processes, and the confirmation or violation of prediction may in turn adjust internal models as well as alter the encoding and retrieval of actual experiences. In addition, on a longer time scale, the interaction between predictive processing and memory may change over the lifespan due to divergences in developmental orientation (progression vs. conservation) and neurocognitive landscape (structural and functional integrity of memory neural systems).

In a series of talks and breakout sessions in POEM, we will discuss and integrate recent theoretical and empirical work related to the aforementioned aspects. We cover behavioral and neuroimaging data as well as contributions from computational modeling. We aim to find links between the different lines of research, identify common principles, and provide a more comprehensive picture of predictive processing as being part of a dynamic memory system across short-term and long-term time scales.