NEURAL AND BEHAVIORAL DYNAMICS OF SOCIAL LEARNING IN MICE

“Copying when uncertain” is a common social learning strategy that helps animals make decisions in complex environments. To study this behavior, we developed a visually guided two-alternative forced-choice task for freely moving mice. After learning the rule during solo training, mice perform joint sessions paired with a partner. Our previous work showed that, on uncertain trials, mice bias their choices toward those of a conspecific. Here, we asked whether this copying strategy depends on specific social cues or whether it generalizes to non-social, artificial cues. To test this, we engineered a high-performance mechanical slider designed to reproduce key features of a mouse partner, including movement dynamics and choice statistics.
Although the slider achieved high task accuracy, mice showed a delayed emergence of copying and lower copying weights compared with sessions paired with a real conspecific. Moreover, copying weights adapted dynamically to partner identity, decreasing with the slider and increasing with a real mouse, highlighting the strong social specificity of this strategy.
To identify the circuit mechanisms underlying this specificity, we focused on the medial prefrontal cortex (mPFC), motivated by whole-brain c-Fos mapping showing elevated mPFC activity during joint versus solo sessions. We used retrograde viral tracing to map upstream inputs to mPFC that may contribute to partner-dependent copying. Finally, to test causal necessity, we silenced candidate circuits using the inhibitory opsin stGtACR2. Preliminary results indicate that mPFC inhibition prolongs decision times. Together, this work aims to dissect the social nature of “copying when uncertain,” linking behavior to the neural circuits.