Yonatan Loewenstein

This paper models procrastination as a self-consistent policy under hyperbolic discounting and shows how task division and breaks can reduce delay.

This study shows that rats can rapidly adapt to new task contingencies using abstract representations in the orbitofrontal cortex, supporting few-shot learning.

The paper investigates the large-scale organization of somatosensory maps in the cortex, finding radial geometry and orthogonality between body representation and hierarchy in S1, with similar patterns in visual and auditory systems.

The paper shows that the structure of a representational map in the auditory cortex is temporarily disrupted by neuron loss but recovers through plasticity of previously unresponsive neurons.

This study examines spatiotemporal dynamics of locomotor decisions in Drosophila and humans, showing that future choices can be predicted through sustained locomotor signatures and highlighting the role of sensory and dopaminergic signaling.

The paper compares the long-term dynamics of idiosyncratic choice bias and feedback-induced bias, finding that feedback-induced bias is transient while idiosyncratic bias remains stable.

The paper identifies sex-specific changes in cholinergic-targeting tRNA fragments in Alzheimer's disease, linking them to cholinergic neuron loss and dementia.

The paper investigates how background odors affect odor detection speed and accuracy in mice, finding that background odors add noise to decision-making, reducing accuracy while increasing speed.

The paper characterizes hippocampal astrocytic domains and their spatial relationships to excitatory and inhibitory neurons using imaging and computational methods.

The paper quantifies somatosensory representation across the human cortex, showing that body-part proportions differ significantly from S1 and suggest a role in cognition.

The paper shows that stable excitatory synaptic connectivity predicts which CA1 pyramidal neurons will become engram neurons after learning.

The paper quantifies idiosyncratic choice biases in decision tasks and shows they emerge from neuronal network dynamics, not experimental context.

This paper challenges the claim that striatal neurons encode action-values by identifying critical confounds in existing studies and failing to detect action-value representations using improved methods.

The paper shows that deviations from the matching law in foraging behavior reflect an optimal strategy using multi-timescale learning to balance bias and variance under uncertainty.