Representational geometry of punishment learning across basolateral amygdala neurons

Longitudinal calcium imaging in 1,147 basolateral amygdala neurons of male mice shows punishment learning reshapes action-linked activity to align with the punisher, action- and contingency-specific.

• Why it matters: The basolateral amygdala is required for punishment learning, but it was unclear how its population activity evolves with learning and how it represents an action’s consequences.
• What they did: Male mice underwent instrumental training, punishment, a choice test, and punishment extinction while 1,147 basolateral amygdala neurons were tracked with longitudinal calcium imaging.
• The result: Punishment learning transformed activity for the punished action, diverging from the unpunished action and aligning with punisher-evoked activity; reward and shock added largely orthogonal population directions, preserving reward components.

AI summary of the abstract · may contain errors · not medical advice · read the paper.