(PS4-C76) Characterizing the Expression of Fear Extinction in Episodic Memory

Aversive learning overgeneralizes across time, contexts, and related stimuli. In contrast, safety learning, which counteracts conditioned fear, produces poorly generalizable, inflexible, and fragile memories. Threat-safety competition in memory is often investigated in Pavlovian conditioning paradigms, relying on implicit conditioned threat responses. Here, we used a hybrid category conditioning and episodic memory experiment (N=41) to investigate how Pavlovian fear and extinction learning are expressed in episodic memory. On day 1, from semantic category items (animals, tools) served as conditioned stimuli paired (CS+) or unpaired (CS-) with shock, followed by extinction learning. Day 2 featured a mnemonic similarity task, including repeats (old-targets), visually similar (similar-lures), or novel (new-foils) items. Threat and safety stimuli showed enhanced recognition memory, but differing pattern completion (generalization) and separation respectively. Similar-lures were rated ‘old’ (generalized) for threat items, but discriminated as ‘new’ for safe items. Pattern separation also predicted less source memory threat-biases (misattributing items to acquisition). These findings indicate that well-known memory systems may constitute novel features of fear and safety learning; represented by generalized fear memories (pattern completion) and restricted or specific safety memories (pattern separation). Recognizing pattern separation’s role in safety learning may facilitate novel insights on the expression of adaptive emotional memory.