Staff Directory

I am interested in combining behavioral testing, electrophysiological and neuroimaging data, and computational modeling tools to learn about human vision, information processing, and behaviors. My current research interest lies in modeling the face-processing system of macaques.

I am interested in understanding the computational mechanisms that allow the brain to perceive and interpret complex stimuli.
 

My current research is focused on how humans process visual information about natural and synthetically-generated faces. I use face-generative models and probabilistic modeling of human perceptual judgements to understand the factors that drive different aspects of face perception such as “humanness”, face similarity and identity discrimination.

I am also interested in how can we build AI systems that are ethical and fair by incorporating an understanding of human cognitive factors and societal issues.

I am originally from Bogotá, Colombia

I am an undergraduate student at Carnegie Mellon University pursuing a double major in Cognitive Science and Human-Computer Interaction, with a minor in Creative Writing. My research interests lie at the intersection of human-centered AI and cognitive science, with a focus on designing AI systems informed by cognitive theory that can reason, communicate, and adapt in ways that better reflect human cognition, emotion, and perception. My previous work includes modeling social worlds from life stories at the CMU Robotics Institute, building an emotion-aware dialogue system for the Furhat robot at Koç University, and extending AI audio generation and automating a preprocessing pipeline at Stanford University.

At the Visual Inference Lab, I am working with Dr. Akihito Maruya on computational modeling of motion textures, studying dynamic, naturalistic non-rigid motions such as flickering fire, lapping ocean waves, drifting smoke, rustling leaves, and swaying people. Our work asks how the visual system summarizes these complex motions into recognizable percepts, and which motion statistics carry perceptual weight.

Outside of research, I enjoy reading, making ceramics, playing the piano, and playing basketball.

I manage the Visual Inference Lab and support Dr. Kriegeskorte in the administrative, operational, and compliance work that helps the lab’s research run effectively. Before joining Columbia, I earned my B.S. in Biomedical Engineering from the University of Arizona and worked in biotech operations, where I built experience in quality systems, process improvement, and research support. I am currently pursuing a Master’s in Applied Mathematics Outside of work, I tutor math, enjoy running and yoga, and love games of all kinds. 

Are we closing the gaps between humans and machines? This question, which is at the intersection of AI, Brain and Cognition (ABC), has significantly guided my academic pursuits. I aim to translate insights from cognitive science and neuroscience to push the boundaries of AI research, and use AI to advance our understanding of human intelligence. Before joining Columbia, I studied representational alignment between humans and machines under the guidance of Professors Thomas Serre and Drew Linsley at Brown University.

I am a Presidential Scholar in Society and Science and Fellow of the Italian Academy working on the neuroscience and philosophy of 3D vision. At the Visual Inference Lab I’ll be developing the themes of my recent Royal Society meeting New Approaches to 3D Vision, which seeks to integrate human and computer vision approaches, as well as my book The Perception and Cognition of Visual Space (Palgrave, 2017). My PhD was from the Centre for Applied Vision Research, City, University of London, and I was also previously part of the DeepFocus team at Meta (Facebook) Reality Labs. Prior to vision science I was a Stipendiary Lecturer in Law at Oxford University and a Teaching Fellow in Philosophy at UCL.

I am primarily interested in probabilistic machine learning with research experience in MCMC methods, variational inference, Bayesian nonparametrics, and Bayesian active learning. In the Kriegeskorte lab, I've focused so far on more flexible Bayesian models for crowdsourcing/classifier combination with state-of-the-art performance on benchmark datasets under two separate novel approaches. More recently I have also been working on making inference-time scaling in LLMs more efficient using probabilistic methods.

contact: [first name][last name][at]gmail[dot]com

I am an undergraduate student at the University of Rochester (class of 27') pursuing a double major of B.S. in Computer Science and a B.S. in Brain and Cognitive Science, as well as a Japanese Minor and a concentration in AI. 

At UofR, I have been doing research with the Samuel Norman-Haignere Auditory Computational Neuroscience lab, where I lead a project to analyse acoustic invariance to environmental sounds, analysing the mechanisms that underlie auditory classification as well as the capabilities of encoding models to replicate them.

My interests lie in the intersections between computer science and cognitive science. I want to be able to understand how intelligent systems arise, what computations and mechanisms underlie them, and how we can interact and understand such systems. This has led me to Computational Neuroscience, with particular interests in NeuroAI, Brain Computer Interfaces, and Neural Decoding. I am interested in perception, especially imagined visual and auditory stimuli. 

For fun I enjoy playing the guitar and making music, and at UofR I am the club president of the Rochester Producers and Musicians club. I also love nature and dogs. 

I want to understand how our brains enable us to see. Opening our eyes gives us an almost instant sense of our surroundings, so visual computations must be rapid. This suggests that the system uses fast feedforward computations to map from retinal images to abstract representations of the scene and the objects. However, understanding the structure of the scene, the relationships among the objects, and their implications requires relating the visual signals to prior knowledge about the world in a deep and highly flexible way. This suggests that vision is also an inference process that involves the active construction of internal models that reflect both prior knowledge and present evidence. In machine learning, these two paradigms of perceptual processing have been somewhat separately explored with feedforward computational models (which still dominate computer vision) and probabilistic generative models (which more cleanly separate the roles of the prior knowledge and the inference algorithm, and promise powerful generalization to new perceptual challenges). The primate brain combines the computational efficiency of the former paradigm with the statistical efficiency of the latter. It appears to seamlessly integrate these two computational paradigms using an algorithm that is yet to be discovered. This algorithm is the central mystery that drives my interest in vision.

My lab uses deep neural networks, a brain-inspired artificial intelligence technology, to build computer models that can see and recognize objects in ways that are similar to biological visual systems. We use a top-down engineering approach to design models to perform complex visual tasks and match human behavioral performance. In addition, the models are constrained, from the bottom up, by neuroscientific data. They must use only neurobiologically plausible dynamic components and must be able to explain the internal image representations and dynamic transformations observed in biological brains with techniques including functional magnetic resonance imaging, magnetoencephalography, and cell-array recordings.

Beyond building computational models of biological vision, my lab works on methods for testing such models with brain and behavioral data. Just like vision must relate complex models of the world to the massive stream of retinal data, computational neuroscience needs to test complex neural network models with increasingly rich measurements of brain activity and behavior that we can now acquire in humans and animals. We are developing exploratory visualization methods for high-dimensional data, as well as confirmatory methods for inferential comparisons among brain-computational models.

More information on the lab's general approach is here.

Kriegeskorte is a Professor of Psychology and Neuroscience at Columbia University. He is an affiliated member of the Department of Electrical Engineering. He is also a Principal Investigator and Director of Cognitive Imaging at the Zuckerman Mind Brain Behavior Institute. Kriegeskorte is a co-founder of the conference “Cognitive Computational Neuroscience”, which had its inaugural meeting in September 2017 at Columbia University. Kriegeskorte received his MA from the University of Cologne in Germany, did his PhD thesis research at Maastricht University in the Netherlands, and worked as a postdoctoral fellow at the University of Minnesota and at the US National Institute of Mental Health. From 2009 to 2017, he was a Programme Leader at the Medical Research Council's Cognition and Brain Sciences Unit at the University of Cambridge, UK.

My research examines the general principles of neural organization that enable fast, efficient visual perception in humans. To uncover these, I build neural network models that faithfully integrate primate neurobiology in order to solve visual tasks. I am more generally interested in the neural bases of sensation, cognition, and movement as a window into general intelligence.

Before joining the lab, I received my B.S. in Cognitive Science from UCLA and worked in Dr. Winrich Freiwald's lab at Rockefeller University as a research assistant.

I'm generally interested in computational neuroscience, philosophy of Mind, and AI safety. I received my B.A. in computer science and philosophy from UNC Chapel Hill. Before that, I briefly pursued a B.S. in physics at USTC. 

I am a postdoc in the lab and employ neuroimaging and deep learning approaches to characterize how intelligent visual systems, whether biological or artificial, can flexibly encode visual feature conjunctions in service of adaptive behavior. I completed my PhD in the Harvard Vision Lab working with Yaoda Xu.

I am interested in using deep neural networks to study how the visual system combines top-down and attention modulations with bottom-up and lateral connectivity to efficiently group visual input into objects. At the Visual Inference Lab, I build computational models of the brain mechanisms of face perception and evaluate them using controversial and adversarial stimuli to drive models and brain responses during face recognition.

I am interested in developing computational tools for understanding cognitive functions, especially how the brain constructs the world model and gives rise to subjective perceptual experiences. At the Visual Inference Lab, I work on computational modeling of visual illusions using deep neural networks.

I received a PhD in Informatics from Kyoto University, working on reconstructing subjective visual experiences from brain activity, supervised by Prof. Yukiyasu Kamitani. I regard the reconstruction technique as an "inner" psychophysical tool and applied it to probe neural representations of visual illusions. Previously, I worked with Prof. Jianfeng Feng and earned a B.S. in Mathematics from Fudan University.

I am a visiting student from Maastricht University studying Cognitive Neuroscience (M.S.). I especially enjoy the study of the emergence of percepts and other lower-order cognitive processes that cross the boundary from the unconscious to the conscious. At the Visual Inference Lab, I am joining Dr. Fan Cheng in her work on visual illusions, specifically using deep neural network modeling to identify visual stream architectures that could explain how the brain gives rise to visual illusions.

I study how resource constraints—space, time, energy, and data—shape intelligence, and how brains and machines build internal models of the world. I received a BA in Philosophy from King’s College London and an MSc in Computer Science from Imperial College London. Before joining Niko's lab, I was an RA in Ilker Yildirim’s group at Yale, where I worked on computational models of dynamic attention.

I am interested in understanding how representations in the brain support computations underlying perception and learning. I believe that deep unsupervised learning can help us uncover the structure of neural and behavioral data to better answer this question. 

Previously I studied computer science and cognitive science at Brown, after spending my freshman year at Middlebury College.

My research focuses on understanding how the brain recognizes faces and its underlying computational mechanisms. At the Visual Inference Lab, I leverage artificial neural networks (ANNs) to simulate diverse face recognition tasks and apply a distinct method—'Artiphysiology'—to manipulate these models and investigate how invariant face recognition emerges at single-cell and population levels. Working closely with experimentalists, I test model predictions against neural data to uncover the key principles driving face perception.

I completed my PhD with Dr. Zaidi at SUNY’s Graduate Center for Vision Research, where I studied psychophysics and computational modeling of 3D shape perception, object rigidity and non-rigidity, and perceptual distortions in amblyopia. This work uncovered computational principles of when and why the human visual system operates sub-optimally, which sparked my interest in how humans learn to see.

My current research focuses on how humans acquire visual representations in an unsupervised manner that enables them to navigate the world. The fovea, which provides the highest resolution in our visual field, spans only about 1° out of the ~220° field of view. Yet our perception feels seamless, as if high resolution were available everywhere. One possibility is that we continuously predict foveal detail from peripheral input, a process that requires knowledge of the statistical structure of natural scenes.

I aim to implement this unsupervised learning paradigm in Vision Transformers, testing the models both on task performance and their ability to predict human perceptual phenomena such as change blindness. I also plan to compare the models’ internal representations with neural data to better understand the link between visual learning in humans and artificial systems.