Hossein Adeli
Associate Research Scientist
Kriegeskorte lab
Zuckerman Institute for Brain Mind Behavior
Columbia University
New York, NY
I work at the intersection of neuroscience and AI (Neuro-AI). We build encoding models using modern machine learning and deep neural networks that predict neural activity to better understand computations in the brain. These models can also serve as “digital twins” on which we perform in silico experiments to reveal the selectivity of different units. We also are building state of the art decoding models to reconstruct perceived stimuli from brain activity. In other work, I study how the visual system combines top-down and attention modulations with bottom-up and lateral connectivity to efficiently group visual input into objects and reveal mechanisms of face perception.
Previously, I was a Senior Postdoctoral Fellow in the Eye Cognition Lab at Stony Brook University psychology department, where I had also recieved my Ph.D. in Cognitive Science.
curriculum vitae 
/
google scholar /
github
ha2366@columbia.edu
hossein.adelijelodar@gmail.com
News - Updates
Sep 2025 - Our work on Transformer brain encoders explain human high-level visual responses was accepted to Neural Information Processing Systems (NeuRIPS 2025) Spotlight (top %3) [preprint][code]
Sep 2025 - Collaboration with Andrew Luo’s group on Meta-Learning an In-Context Transformer Model of Human Higher Visual Cortex was accepted to Neural Information Processing Systems (NeuRIPS 2025) [preprint]
Sep 2025 - Our work on In Silico Mapping of Visual Categorical Selectivity Across the Whole Brain was accepted to Neural Information Processing Systems (NeuRIPS 2025). [earlier preprint]
Sep 2025 - Gave gures lectures introducing the students in Parson School of design to Neuro-AI [Watch here]
Aug 2025 - Presented our work on at NeuroAdapter: Visual Reconstruction with Masked Brain Representation at the Cognitive Computational Neuroscience meeting, Amsterdam [Short preprint]
May 2025 - Gave a talk on Capturing the Representational Dynamics of Face Perception in Deep Recurrent Neural Networks at MODVIS 2025: Computational and Mathematical Models in Vision** workshop at Vision Sciences society meeting
April 2025 - Gave a talk on Flexible relational neural encoder architecture explains human high-level visual responses at Department of Systems and Computational Biology, Albert Einstein College of Medicine.
September 2024 - Gave a talk on Recurrent models optimized for face recognition exhibit representational dynamics resembling the primate brain at SciFest at the Zuckerman Mind Brain Behavior Institute. [Short preprint]
August 2024 - Gave a talk on Predicting brain activity using Transformers at Center for Theoretical Neuroscience, Columbia University.
June 2024 - Our work on The attentive reconstruction of objects facilitates robust object recognition was published in PLOS Computational Biology [link]
June 2024 - Gave a presentation at Amirkabir (Tehran Polytechnic) university in Farsi on our work on predicting brain activity with transformers. [talk on youtube]
May 2024 - Presented our work on object grouping at VSS 2024 and at our symposium Using deep networks to re-imagine object-based attention and perception.
April 2024 - Developed a tutorial for the neuromatch neuroai course on Computation as transformation of representational geometries. Check it out (here)
Feb 2024 - Our symposium submission on Using deep networks to re-imagine object-based attention and perception is accepted for the 2024 Vision Sciences Society meeting (VSS)
Nov 2023 - Gave a talk on our work on predicting brain activity with transformers at Latin American Workshop on Computational Neuroscience (LAWCN)
Oct 2023 - Gave a talk on our work on predicting brain activity with transformers at NeuroAI Montreal
Aug 2023 - Gave a talk on our work on predicting fMRI activity using encoder-decoder transformers at CCN2023 [talk on youtube]
Aug 2023 - Gave a poster presentation on our work on predicting dynamics of object grouping in humans at CCN2023 [poster]
July 2023 - Our submission to algonaust achieved 2nd place [Algonauts leader board][report][code]
June 2023 - New preprint on predicting dynamics of object grouping in humans using self-supervised transformers [paper][code]
May 2023 - Gave a poster presentation on our work on predicting dynamics of object grouping in humans at VSS2023
May 2023 - Released our dataset on dynamics of object grouping in humans [data]
May 2023 - Our work on A brain-inspired object-based attention network for multiobject recognition and visual reasoning is out in Journal of Vision [paper][preprint][code]