John Pearson

This paper introduces a dual flow approach for modeling neural data, producing more interpretable dynamics and better reconstructions than existing methods.

The paper investigates how neurons in the frontal eye field encode priors for visual stability judgments, finding that they support flexible categorization rather than Bayesian inference.

This study identifies specific cortico-basal ganglia synapses critical for rapid vocal learning in juvenile zebra finches and characterizes the timescale of consolidation.

The paper shows that motor cortical circuits integrate distorted proprioceptive feedback by separating activity subspaces to maintain task-relevant movement output.

This study investigates how 15 emotional states are represented in the brain using fMRI and finds that categorical labels better correlate with brain activity than dimensional ratings.

This paper models how juvenile zebra finches learn song through self-guided reinforcement learning by developing a neural circuit that generates internal error signals via predictive cancellation.

The paper examines how light adaptation affects the spatiotemporal receptive field structures of retinal ganglion cell types in rats, finding that their relative organization is largely preserved.

The paper uses brain imaging to investigate how narrative and utility processes contribute to juror decision making by analyzing patterns of brain activation during evidence evaluation.

The paper investigates whether active visual perception during saccades follows Bayesian principles, finding that subjects use Bayesian models for continuous judgments and discriminative models for categorical judgments based on uncertainty sources.

The paper uses efficient coding theory to show how retinal mosaics and cell types emerge with varying channel capacity and noise levels.

The paper identifies odorant ligands for over 500 mouse olfactory receptors and shows that molecular properties of odorants predict receptor responses and selectivity.

This paper introduces a generative model for task-based fMRI analysis that produces group and single-subject effect maps with improved flexibility and interpretability compared to standard methods.

The paper uses efficient coding theory to determine the optimal spatial arrangement of ON and OFF receptive fields in the retina based on noise and stimulus statistics.

improv enables real-time adaptive neuroscience experiments by integrating modeling, data collection, and analysis across organisms and data types.

The paper identifies two forebrain inhibitory neuron populations that trigger or suppress vocalization by modulating PAG-USV neurons through distinct synaptic mechanisms.

This paper shows that both monkeys and humans use causal inference to determine the number and location of auditory and visual stimuli, with similar behavioral responses.

This paper uses Bayesian nonparametric models to analyze dynamic strategic interactions in a competitive game, quantifying agent behavior through Gaussian Processes and reinforcement learning frameworks.

An automated pipeline maps head-mounted gaze onto a fixed 2D stimulus without tethering participants, plus a task built to compare calibration accuracy and precision across 3 commercial headsets.