Gordon Pipa

This study shows that incongruent human agents in a virtual environment increase gaze variability and improve spatial recall through more flexible visual exploration.

The paper shows spike synchrony reliably detects stimulus familiarity while firing rate detects stimulus class, demonstrating complementary coding mechanisms in neural computation.

This paper introduces a spiking neural network model that decodes familiarity from temporal and frequency features of spike trains, outperforming LSTM in continual familiarity detection.

This study examines how human agents affect spatial navigation and knowledge acquisition in a virtual environment, finding that agents influence local attention and navigation accuracy.

The paper demonstrates that spike synchrony in V1 reflects the Gestalt structure of stimuli, suggesting it serves as a mechanism for estimating prior probability of incoming visual input.

This paper proposes a model of neural computation using dendritic plateau potentials to process spike sequences across multiple time-scales, showing how they enable robust detection and processing of sequential patterns.

This study confirms the link between 2D:4D ratio and spatial performance in rats and humans, and identifies brain correlates in humans.

This paper introduces a model using recurrent neural networks to encode and decode dynamic sensory signals, demonstrating its ability to recognize complex hierarchical structures in birdsongs.