Main Menu (Mobile)- Block
Main Menu - Block
Current Research:
Learning general principles from single examples is a hallmark of higher cognitive functions in humans and animals, and a missing property in current AI systems.
In this project, we want to explore several questions related to this remarkable property of intelligent systems:
- which neural properties are used for generalization
- how the brain constructs good features for learning
- what algorithms does the brain use to identify relevant features to be learned in the context of a task
- which circuit changes and computations are supporting this property
Biography
Miguel Nunez has been always motivated by the complexity and interdisciplinary nature that requires understanding the human body, especially the brain, following such motivation, he received a bachelor’s degree from Monterrey Institute of Technology in biomedical engineering, and after several projects in the neuro-engineering, data analysis and machine learning fields guided by Alejandro García-González and Rita Q Fuentes-Aguilar, he pursued a Ph.D. in neuroscience at Guadalajara University at Medina-Ceja lab, in which using a combination of computational modeling with electrophysiological recordings, he described a novel interaction between different regions of the hippocampal formation related with a bio-marker of epilepsy called fast ripple and its possible impact in cognition and memory.
Realizing the complex data analysis problems the brain presents, he finished in parallel to his Ph.D., a program in statistics, data science, and machine learning at the Massachusetts Institute of Technology, where he developed a special interest in the intersection between biological and artificial intelligence, and how the coordinated activity of biological or artificial neurons give rise to perception, decision, and action.
Passionate about the intersection between fields, he's currently a postdoc at Marius Pachitariu lab, the lab straddles the fields of mechanistic cognitive science, computation, and theory at Janelia Research Campus, studying complex behaviors in mice by recording from populations of up to 50,000 neurons and using machine learning techniques, to analyze the structure of those recordings, and theory, to describe and understand the algorithms used by the brain in the tasks.
Miguel’s current research focuses on generalization, a key property of intelligence that allows biological intelligence to arrive at general principles from limited information, being such property a fundamental piece missing in current artificial intelligence systems, this project spans multiple research directions, some of them in collaboration with Carsen Stringer lab.
In addition, he is interested in studying changes in neural representations due to learning, the use of memory as a useful prior for generalization, and the use of brain-computer interfaces as a probe tool for learning.
His work and passion for data science, neuroscience, and intelligence science were awarded in Mexico by the “Innovation, Science, and Technology” Award in 2020.