Ashok Litwin-Kumar

This paper develops a theory of multi-tasking in neural circuits, distinguishing between sequential and simultaneous task performance and introduces a model of task barcodes to parameterize task selectivity.

The paper develops a theoretical model to explain multi-task computation in neural networks, showing how task selection can overcome interference between different task dynamics.

The paper introduces a graph embedding algorithm to identify symmetries in neural connectomes, demonstrating its ability to detect rotational, translational, and toroidal symmetries in Drosophila and synthetic grid cell connectomes.

The paper develops a framework to predict neural network activity using connectome-constrained recurrent networks and shows that targeted neuron recordings can resolve ambiguities in network dynamics.

The paper shows that inducing backward movement in Drosophila engages dopaminergic punishment and supports aversive memories, explaining the avoidance paradox.

The paper shows that experience reorganizes inhibitory neuron connectivity in the mouse olfactory cortex, increasing network response dimensionality and discriminability of familiar stimuli.

The paper shows that striatal dopamine and acetylcholine encode trajectory errors during navigation, with distinct spatial and temporal patterns, and links these signals to learning and behavior.

The paper identifies visual inputs to Kenyon cells in Drosophila mushroom bodies, showing sparse, distributed, and combinatorial coding similar to olfactory coding but with distinct input repertoires.

The paper compares connectivity patterns in the Drosophila mushroom body across species to show how evolutionary changes in neuronal connectivity relate to chemical ecology.

The paper identifies optimal corticothalamic connectivity structures that support learning in thalamocortical loops for motor control and working memory tasks.

The paper resolves a contradiction between striatal reinforcement learning models and SPN synaptic plasticity by showing that iSPN plasticity hinders learning but can be reversed by simultaneous activation of dSPN and iSPN.

This paper proposes a model of recall-gated consolidation that enhances memory storage by prioritizing reliable synaptic changes, showing benefits in noisy environments and explaining behavioral learning phenomena.

This study determines the optimal granule cell representation for cerebellar learning tasks beyond random stimulus discrimination, finding that dense representations are more optimal for continuous input-output transformations.

The paper develops a general theory of cerebellum-like structures and their afferent pathways, predicting functional roles and reconciling existing theories.

The paper identifies a feedforward circuit between dopamine subsystems in Drosophila that is essential for second-order conditioning, mediated by a key interneuron.

The paper shows that mushroom body input connections in Drosophila form without sensory activity, suggesting genetic programming over experience-based wiring.

The paper develops a modeling approach to understand how heterogeneous dopamine signaling in Drosophila mushroom bodies coordinates plasticity during learning tasks.

The paper maps the developmental origin of interneurons in Drosophila and shows how lineage and temporal identity shape synaptic connectivity and circuit formation.

The paper maps the synaptic connectivity of the Drosophila mushroom body and identifies new circuit components and structures in sensory input processing and output modulation.

The paper constructs a Drosophila larval connectome linking proprioceptors, premotor neurons, and motor neurons to explain how distinct motor behaviors arise from shared neural circuits.

The paper shows nitric oxide acts as a cotransmitter in dopaminergic neurons to modulate memory dynamics in Drosophila.

The paper maps the synaptic connectome of the Drosophila larva mushroom body and identifies feedback pathways that enable modulatory neurons to compute teaching signals for reinforcement learning.

The paper investigates how context influences olfactory decision-making by examining the role of the premotor area ALM in a delayed match-to-sample task.

The paper reconstructs the complete synaptic connectome of the Drosophila larval mushroom body, revealing novel circuit structures and input integration patterns in Kenyon cells.