​Psychology Seminar | The dynamics of memory—decoding brain activity with machine learning

Even though encoding and retrieval processes are both critical determinants of performance in memory tests, only their joint effects are observable in overt behaviour. This has led to contentious debates about the nature of the signal elicited by recognition memory probes and about the relative contributions of encoding and retrieval processes in interactions between semantics and episodic memory systems. Using machine learning techniques to quantify relevant signals in brain activity as they unfold during engagement in memory tasks I will address these controversies. Specifically, I will present evidence for a single-route account of recognition memory that is compatible with contributions from familiarity and recollection signals, but relies on a unitary evidence signal that integrates all available evidence. I will also particularly implicate retrieval (rather than encoding) processes in the categorical organization of episodic memories.