Long-term memory interference is resolved via repulsion and precision along diagnostic memory dimensions

Abstract

When memories share similar features, this can lead to interference, and ultimately forgetting. With experience, however, interference can be resolved. This raises the important question of how memories change, with experience, to minimize interference. Intuitively, interference might be minimized by increasing the precision and accuracy of memories. However, recent evidence suggests a potentially adaptive role for memory distortions. Namely, similarity can trigger exaggerations of subtle differences between memories (repulsion). Here, we tested whether repulsion specifically occurs on feature dimensions along which memories compete and whether repulsion is predictive of reduced memory interference. To test these ideas, we developed synthetic faces in a two-dimensional face space (affect and gender). This allowed us to precisely manipulate similarity between faces and the feature dimension along which faces differed. In three experiments, participants learned to associate faces with unique cue words. Associative memory tests confirmed that when faces were similar (face pairmates), this produced interference. Using a continuous face reconstruction task, we found two changes in face memory that preferentially occurred along the feature dimension that was “diagnostic” of the difference between face pairmates: (1) there was a bias to remember pairmates with exaggerated differences (repulsion) and (2) there was an increase in the precision of feature memory. Critically, repulsion and precision were each associated with reduced associative memory interference, but these were statistically dissociable contributions. Collectively, our findings reveal that similarity between memories triggers dissociable, experience-dependent changes that serve an adaptive role in reducing interference.

Object feature memory is distorted by category structure

“…We extend this literature by showing how different kinds of features exhibit varying susceptibility to memory distortions depending on their role in a category. Although there has been limited prior work examining feature-level memory distortions in category learning, one study found that participants misremembered pairs of faces as less similar to one another (repulsion) along the facial feature dimension most diagnostic for differentiating the two faces (Drascher & Kuhl, 2022). This finding is consonant with the findings we report here for unique features, where information that is diagnostic of an item showed reduced attraction (though not repulsion).…”

“…Related distortions in perception have also been long-documented via the phenomenon of categorical perception (Dubova & Goldstone, 2021;Folstein et al, 2015;Goldstone, 1994), where features near a category boundary are perceived as more distinct. Our work differs from this literature, as well as Drascher & Kuhl (2022), in showing how features relating to the internal structure of a category, as opposed to discrimination boundaries, are distorted, and demonstrating effects in an attribute (color) that does not itself carry information about feature roles (whether a feature is unique or shared) in the category.…”

“…However, while shared features showed more attraction than unique ones, we still saw strong overall attraction biases for both feature types in our paradigm. In other words, we never observed true repulsion (Chanales et al, 2017(Chanales et al, , 2021Chunharas et al, 2018;Drascher & Kuhl, 2022;Favila et al, 2016;Wanjia et al, 2021;Zhao et al, 2021), where an item's color would be misremembered as more distinctive from related information than it actually was. One possibility is that this extreme form of separation for unique features would have emerged if the training period in our study was extended longer, as both experiments showed decreasing attraction for unique features over learning.…”

Behavioral representational similarity analysis reveals how episodic learning is influenced by and reshapes semantic memory

“…This is precisely what we found in our analyses of lure representationstesting exerted the biggest impact on moderatestrength lures, which were significantly repelled away from cue words, relative to weak/non-lure pairings. This effect is not only consistent with the predictions of NMPH, but also with an emerging body of work showing that competition adaptively distorts and repels overlapping episodic representations so they become less similar 46,64,65 .…”

Episodic learning-induced representational change is both influenced by and re-sculpts semantic memory

“…Prior experimental work has demonstrated that representations of highly similar episodic memories are both reactivated during memory retrieval 68,69 . Competition adaptively distorts overlapping episodic representations so they become less similar; the amount that overlapping representations are repelled is predictive of subsequent memory 44,70,71 . We showed that potential semantic lures are pushed away from the cue in representational space only for semantically related pairs, suggesting that preexisting connections between cue and target can be used as retrieval cues, but that additional shared concepts must be repelled for the connections to be used successfully.…”