Patrick Marino, Ph.D.
I develop machine learning systems for neural interfaces to help people with motor impairments regain movement. I'm currently a Senior Research Scientist at Phantom Neuro. Previously, I interned at Neuralink. In my doctoral research, I used brain-computer interfaces to study the neural control of movement with Aaron Batista, Byron Yu, and Steve Chase.
Projects and publications
A posture subspace in primary motor cortex
Neuron (2025)
To generate movements, the brain must combine information about movement goal and body posture. The motor cortex is a key node for the convergence of these information streams. How are posture and goal signals organized within M1’s activity to permit the flexible generation of movement commands? To answer this question, we recorded M1 activity while monkeys performed a variety of tasks with the forearm in a range of postures. We found that posture- and goal-related components of neural population activity were separable and resided in nearly orthogonal subspaces. The posture subspace was stable across tasks. Within each task, neural trajectories for each goal had similar shapes across postures. Our results reveal a simpler organization of posture signals in M1 than previously recognized. The compartmentalization of posture and goal signals might allow the two to be flexibly combined in the service of our broad repertoire of actions.
A neural basis of choking under pressure
Neuron (2024); PNAS (2019)
Incentives tend to drive improvements in performance. But when incentives get too high, we can "choke under pressure" and underperform right when it matters most. What neural processes might lead to choking under pressure? We studied rhesus monkeys performing a challenging reaching task in which they underperformed when an unusually large "jackpot" reward was at stake, and we sought a neural mechanism that might result in that underperformance. We found that increases in reward drive neural activity during movement preparation into, and then past, a zone of optimal performance. We conclude that neural signals of reward and motor preparation interact in the motor cortex (MC) in a manner that can explain why we choke under pressure.
Dynamical constraints on neural population activity
Nature neuroscience (2025)
The manner in which neural activity unfolds over time is thought to be central to sensory, motor and cognitive functions in the brain. Network models have long posited that the brain’s computations involve time courses of activity that are shaped by the underlying network. A prediction from this view is that the activity time courses should be difficult to violate. We leveraged a brain–computer interface to challenge monkeys to violate the naturally occurring time courses of neural population activity that we observed in the motor cortex. This included challenging animals to traverse the natural time course of neural activity in a time-reversed manner. Animals were unable to violate the natural time courses of neural activity when directly challenged to do so. These results provide empirical support for the view that activity time courses observed in the brain indeed reflect the underlying network-level computational mechanisms that they are believed to implement.
Towards a robust movement onset decoder for ECoG-based
brain-computer interfaces
Master's Thesis, University of Texas at Austin (2018)
Electrocorticography (ECoG) has shown promise as a recording modality for brain-computer interface applications. However, further work is needed to make ECoG-driven BCIs safe and reliable. One way to improve their safety is to prevent unwanted robotic motion until detecting the user's intent to begin a new movement. We used features of neural activity known to accompany movement onset to create a decoder which detected whether or not the user was trying to begin a movement. Our decoder had a low false-positive rate while maintaining sufficient sensitivity for online control, suggesting a path forward for eliminating unwanted robotic motion in ECoG BCIs.
Skills
Data science and machine learning
I use machine learning and statistics to make sense of complex neural and physiological data, ultimately gaining insights about how the brain controls our movements. I'm proficient in standard methods for regression, classification, and dimensionality reduction.
Science communication
I tell stories with data, communicating complicated ideas in simple ways to help others understand them.
Signal processing
Before analyzing data, I preprocess raw signals using standard filtering, smoothing, and artifact removal techniques.
Experimental design and data collection
I design and conduct experiments which carefully remove confounds to reveal relationships between variables of interest.
Mechanical design and fabrication
I design and fabricate mechanical and mechatronic devices for use in experiments, utilizing CAD, statics, dynamics, and robotics in the process.
Experience and Education
Experience
Senior Research Scientist | Phantom Neuro (2024-present)
Neuroengineer Intern | Neuralink (2023-2024)
Doctoral student | University of Pittsburgh (2018-present)
Advised by Aaron Batista and Byron Yu
The influence of posture and reward signals on motor cortical population activity
Master's student | University of Texas at Austin (2015-2018)
Advised by Ashish Deshpande
Towards a robust reach onset decoder for electrocorticography-based BCI’s
REU Summer Research Assistant | Georgia Institute of Technology (2014)
Advised by Patricio Vela
A low‐cost, wheelchair‐mounted robotic arm to assist paralyzed patients
Education
University of Pittsburgh, Ph.D. in Bioengineering (2024)
University of Texas at Austin, M.S.E. in Mechanical Engineering (2018)
University of Notre Dame, B.S. in Physics and B.S. in Mechanical Engineering (2015)
Contact
You can reach me via email at pmarino162@gmail.com
Also check out my LinkedIn !