Cengiz Pehlevan

This paper introduces a network model inspired by the insect fan-shaped body that guides an agent's movement through memory and control loops.

This paper proposes a model of associative learning in cerebellar cortex ensembles using heterogeneous plasticity, showing improved pattern recognition performance with ensemble size.

This study explores how structured and disordered connectivity in recurrent neural networks affects population dynamics and task performance, finding that minimal restructuring can achieve task-relevant representations.

This study shows that hippocampal reward representations become predictive of reward over time as mice learn a reward-based task.

This paper introduces a model for olfactory compressed sensing that simultaneously detects odor presence and estimates concentrations, showing it can be implemented in a biologically plausible circuit.

This paper develops a recurrent neural network-based reinforcement learning framework to model decision-making deficits in schizophrenia, revealing suboptimal belief updating linked to neural dysfunction.

This paper develops a normative model of place field reorganization during reward maximization, showing it unifies navigation phenomena and improves policy learning.

The paper shows that odor responses in the olfactory cortex of mice are highly matched between hemispheres, allowing decoding of odor identity and laterality through Hebbian plasticity.

The paper develops a theoretical model of Hebbian plasticity that predicts sequence replay speed and explains how tutor signals influence memory formation and retrieval.

The paper shows that partial observation of neural circuits can lead to mechanistic mismatches in data-constrained models of neural dynamics.

The paper proposes a mechanistic model of the Drosophila larval olfactory circuit that explains synaptic organization and function through Hebbian plasticity and activity correlations.

This paper proposes a rate-based Poisson compressed sensing model for the olfactory bulb that accurately detects odors within a single sniff.

The paper studies how C. elegans chemosensory neurons encode olfactory information through ensemble activity and natural stimuli.

This study shows how population codes shape inductive bias for efficient learning, demonstrating that biological codes achieve lower activity while maintaining task-specific efficiency.

This paper shows how natural gradient methods can enable fast sampling in spiking neural networks by leveraging population geometry for probabilistic inference.

The paper investigates representational drift in neural networks, showing that noisy synaptic updates cause coordinated random walks in receptive fields while maintaining population code stability.

This study shows how food deprivation alters olfactory behavior in Drosophila larvae by changing neural processing in the antennal lobe through state-dependent circuit dynamics.

The paper analyzes trial-to-trial timing variability in synfire chains, identifying local, global, and jitter components and validating findings in zebra finch models.

The paper shows that similarity-preserving neural networks develop localized receptive fields that tile manifolds, mirroring brain representations like place cells.

The paper investigates how neural networks can implement flexible motor timing in birdsong and finds that feedforward networks better support this flexibility than randomly connected recurrent networks.

The paper proposes that the PN-KC network in insect olfaction implements an online clustering algorithm derived from the k-means cost function, explaining key features of the circuit.