Multimodal temporal modeling of behavioral reorganization across threat learning and extinction in rats

In 29 male rats, a hidden Markov model identified recurring behavioral configurations across threat learning and extinction, with eight configurations predicting best and separating press-active from vocal-active and low-output patterns.

• Why it matters: Threat learning disrupts a previously learned activity, but it is unclear whether later engagement returns as defensive responding declines, and how behavior reorganizes over time.
• What they did: Rats underwent conditioning, extinction, delayed retesting, and context changes while freezing, 22-kHz/50-kHz calls, and food-reinforced lever responses were tracked across 45,300 five-second windows; temporal order was modeled with a hidden Markov approach.
• The result: In the full extinction group (n=9), the estimated time share in press-active patterns increased on average, while cue-free context controls (n=8) showed decreases; configuration summaries did not add consistent predictive value for later responding or whole-recording fecal-pellet counts.

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