Roberto A. Gulli

New York City, NY

The brain’s episodic memory network is essential for learning from our experiences. When it fails, the results are devastating cognitive disorders like Alzheimer’s disease. Yet, nearly 50 years after its initial description, how the key parts of this network actually compute remains largely unknown.

This knowledge gap fundamentally limits our ability to treat these diseases. The Gulli Lab was established to address this gap.

Our research program has two complementary aims:

To define the computational principles that allow the episodic memory network to transform experiences into memories that guide behavior.

To apply this knowledge to help design and optimize therapies for disorders of learning and memory.

To do this, we combine immersive virtual reality, technologies capable of recording from thousands of neurons simultaneously across the brain, neurmodulation, and computational modelling.

Explore our projects to learn about our approach, and our team and mentorship pages to see how you can join us.

News

Dec 1, 2025	The lab’s first NIH grant has been funded! This R00 grant will support our efforts to understand how neuromodulation of the hippocampus and connected brain regions influences learning and memory. Looking forward to an exciting journey ahead!
Nov 15, 2025	Rob will host a Meet-the-Experts session on “Designing and Implementing Large-Scale Neural Recording Systems” at the Society for Neuroscience (SfN) 2025 conference. If you’re attending SfN, please stop by the IMEC booth to discuss the latest advancements in neural recording technologies and their applications in neuroscience research. Get a sneak peek here.
Sep 1, 2025	The Gulli Lab has opened! Check out our projects page to learn more about our research program and approach. Check out our mentorship and team pages to learn more about joining us.

Selected Publications

Studies of Hippocampal Function in Non-Human Primates

Roberto A. Gulli, and Julio C. Martinez-Trujillo

Encyclopedia of the Human Brain, 2nd Edition 2024

Abstract WEB PDF

The hippocampus is a phylogenetically ancient brain structure that has been shown to be critical for spatial navigation and memory. Decades of research have uncovered neurophysiological correlates of each function in the activity of hippocampal neurons, but debate continues over the primacy of each one. This debate exists, at least in part, because navigational and non-spatial mnemonic signals have been difficult to simultaneously observe and disentangle in lesion and electrophysiology studies. Together, the work reviewed in this chapter shows that some, but not all predictions of spatial and mnemonic theories of hippocampal function are corroborated in monkeys performing tasks that allow for precise measurement and parameterization of behaviour during electrophysiological experiments. The points of divergence from established dogmas may have important implications for neuropsychological and computational theories of hippocampal function across species.
Distinct neural codes in primate hippocampus and lateral prefrontal cortex during associative learning in virtual environments

Benjamin W. Corrigan, Roberto A. Gulli, Guillaume Doucet, Megan Roussy, Rogelio Luna, Kartik S. Pradeepan, Adam J. Sachs, and Julio C. Martinez-Trujillo

Neuron 2022

Abstract WEB PDF

The hippocampus (HPC) and the lateral prefrontal cortex (LPFC) are two cortical areas of the primate brain deemed essential to cognition. Here, we hypothesized that the codes mediating neuronal communication in the HPC and LPFC microcircuits have distinctively evolved to serve plasticity and memory function at different spatiotemporal scales. We used a virtual reality task in which animals selected one of the two targets in the arms of the maze, according to a learned context-color rule. Our results show that during associative learning, HPC principal cells concentrate spikes in bursts, enabling temporal summation and fast synaptic plasticity in small populations of neurons and ultimately facilitating rapid encoding of associative memories. On the other hand, layer II/III LPFC pyramidal cells fire spikes more sparsely distributed over time. The latter would facilitate broadcasting of signals loaded in short-term memory across neuronal populations without necessarily triggering fast synaptic plasticity.
Context-dependent representations of objects and space in the primate hippocampus during virtual navigation

Roberto A. Gulli, Lyndon R. Duong, B W Corrigan, Guillaume Doucet, Sylvain Williams, Stefano Fusi, and Julio C. Martinez-Trujillo

Nature Neuroscience 2020

Abstract WEB PDF

The hippocampus is implicated in associative memory and spatial navigation. To investigate how these functions are mixed in the hippocampus, we recorded from single hippocampal neurons in macaque monkeys navigating a virtual maze during a foraging task and a context–object associative memory task. During both tasks, single neurons encoded information about spatial position; a linear classifier also decoded position. However, the population code for space did not generalize across tasks, particularly where stimuli relevant to the associative memory task appeared. Single-neuron and population-level analyses revealed that cross-task changes were due to selectivity for nonspatial features of the associative memory task when they were visually available (perceptual coding) and following their disappearance (mnemonic coding). Our results show that neurons in the primate hippocampus nonlinearly mix information about space and nonspatial elements of the environment in a task-dependent manner; this efficient code flexibly represents unique perceptual experiences and correspondent memories.
Cross-species 3D virtual reality toolbox for visual and cognitive experiments

Guillaume Doucet, Roberto A. Gulli, and Julio C. Martinez-Trujillo

Journal of Neuroscience Methods 2016

Abstract WEB PDF

Background Although simplified visual stimuli, such as dots or gratings presented on homogeneous backgrounds, provide strict control over the stimulus parameters during visual experiments, they fail to approximate visual stimulation in natural conditions. Adoption of virtual reality (VR) in neuroscience research has been proposed to circumvent this problem, by combining strict control of experimental variables and behavioral monitoring within complex and realistic environments. New method We have created a VR toolbox that maximizes experimental flexibility while minimizing implementation costs. A free VR engine (Unreal 3) has been customized to interface with any control software via text commands, allowing seamless introduction into pre-existing laboratory data acquisition frameworks. Furthermore, control functions are provided for the two most common programming languages used in visual neuroscience: Matlab and Python. Results The toolbox offers milliseconds time resolution necessary for electrophysiological recordings and is flexible enough to support cross-species usage across a wide range of paradigms. Comparison with existing methods Unlike previously proposed VR solutions whose implementation is complex and time-consuming, our toolbox requires minimal customization or technical expertise to interface with pre-existing data acquisition frameworks as it relies on already familiar programming environments. Moreover, as it is compatible with a variety of display and input devices, identical VR testing paradigms can be used across species, from rodents to humans. Conclusions This toolbox facilitates the addition of VR capabilities to any laboratory without perturbing pre-existing data acquisition frameworks, or requiring any major hardware changes.