Test dissociable causal roles of medial frontal cortex neural codes in the ABCD task

Determine whether, in rodents performing the ABCD sequential navigation task, perturbing distinct medial frontal cortex neural populations leads to the predicted dissociable deficits: specifically, whether disrupting neurons encoding beliefs about the current task phase impairs inference of the next goal, and whether disrupting neurons encoding the grounding likelihood impairs flexible mapping between task phases and spatial locations.

Background

The authors map model-derived neural codes to distinct computational roles: goal-progress coding to goal expectations, goal-identity coding to beliefs about the current task phase, and goal-identity-with-location coding to the grounding likelihood. They predict that perturbations to these populations should have dissociable behavioral consequences and explicitly note that such predictions remain to be tested.

References

Another critical prediction concerns the functional role of the neural activity patterns reported in the medial frontal cortex of rodents during the ABCD task \citep{el2024cellular}. Our model maps distinct populations of neurons—tuned to goal progress, goal identity, and goal identity in conjunction with spatial location—onto distinct computational processes, namely: goal expectations, beliefs about the current task phase, and the grounding likelihood. As a result, perturbing these neurons should have dissociable effects. For example, disrupting neurons that encode beliefs about the current task phase should impair the animal’s ability to correctly infer its next goal, whereas disrupting neurons that encode the grounding likelihood should impair the animal’s ability to flexibly link task phases to spatial representations. These predictions remain to be tested in future experiments.

— Schema-based active inference supports rapid generalization of experience and frontal cortical coding of abstract structure  (2601.18946 - Maele et al., 26 Jan 2026) in Discussion