Back

Emotion increases the similarity between neural representations of pictures, but not their perceived similarity.

Riberto, M.; Paz, R.; Pobric, G.; Talmi, D.

2021-06-22 neuroscience
10.1101/2021.06.21.449164 bioRxiv
Show abstract

Stimuli that evoke the same feelings can nevertheless look different and have different semantic meanings. Although we know much about the neural representation of emotion, the neural underpinnings that govern judgements of emotional similarity are unknown. One possibility is that the same brain regions will represent similarity between emotional and neutral stimuli, perhaps with different strengths. Alternatively, emotional similarity could be coded in separate regions, possibly those known to express emotional valence and arousal preferentially. In behaviour, the extent to which people consider similarity along the emotional dimension when they evaluate the overall similarity between stimuli has never been investigated. While the emotional features of stimuli may dominate explicit ratings of similarity, it is also possible that people neglect the emotional dimension as irrelevant. We contrasted these hypotheses with two measures of similarity and two different databases of complex negative and neutral pictures, the second of which afforded exquisite control over semantic and visual attributes. Emotion increased neural similarity in a set of regions that represented both emotional and neutral stimuli, including the inferior temporal cortex, the fusiform face area, and the precuneus. Emotion also increased neural similarity in early visual cortex, anterior insula and dorsal anterior cingulate cortex, despite no increase in BOLD-signal amplitudes in these regions. Despite the stronger neural similarity between emotional stimuli, participants rated pictures taken from two distinct emotional categories as equally similar. These results contribute to our understanding of how emotion is represented within a general conceptual workspace.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.