Empathy and prosocial behavior are fundamental to social life, and understanding their neural mechanisms is a central goal of social neuroscience. Animal models offer unique opportunities to investigate these mechanisms at the cellular level, but realizing this potential requires rigorous methods. In this commentary, we raise a set of methodological concerns about an article published in Nature by Zhang et al. (2024), who claim to have identified a cortical mechanism in mice for helping conspecifics in pain. We identify several issues that undermine the authors’ conclusions: inconsistent criteria for interpreting neural overlap as evidence for distinct versus shared representations; decoder results that are at odds with the claimed representational structure; potential biases arising from fixed experimental order; and a mismatch between region-wide experimental manipulations and population-specific mechanistic claims. We situate these concerns within a broader methodological challenge in neuroscience: the difficulty of inferring functional specificity from limited experimental contrasts. Wherever feasible, we offer concrete suggestions for future research. By calling attention to these methodological challenges, this commentary aims to strengthen standards for mechanistic inference in social and systems neuroscience.
Iannetti et al. (2026) studied this question.