A mathematical model built around the assumption that the desire to maintain internal homeostasis drives the behavior of animals, by affecting their learning processes, can explain many real-world behaviors, including some that might otherwise appear irrational.

Real-time imaging of neural activity in behaving mice reveals that the two canonical types of neurons in the olfactory tubercle carry distinct information about learned odor cues, with D1 neurons representing stimulus valence and D2 neurons mainly representing stimulus identity.

Orbital frontal cortex projection neuron activity is necessary to update state-dependent value change to control model-based behavior.

Neuron ensembles in the medial prefrontal cortex gradually develop codes for relevant, latent variables common across multiple experiences while – apparently independently – losing information about irrelevant, contextual variables unique to each experience.

For mice, knowing when it is time to feed is dependent on the neurotransmitter dopamine and the D1R receptor of neurons in the dorsal striatum.

Recordings from serotonin-producing neurons in the brain reveal that these neurons are highly activated by sudden changes in previously familiar environments, potentially explaining why serotonin is important for learning to adapt to such changes.

Hunger neurons promote feeding by triggering a long-lasting potentiation of the rewarding properties of food.

In adolescent rats, whose actions are more impulsive, neuronal striatal activity that precedes self-initiated action sequences has a steeper modulation by waiting time compared to the modulation found in adults.

Cross-modal stimuli are represented in the primary gustatory cortex according to their sensory identity, associability and predictive value.

Two non-overlapping neural circuits that involve the lamina terminalis and arcuate nucleus of hypothalamus independently drive feedforward anticipatory suppression and activation of vasopressin neurons by drinking and feeding, respectively.