About

Professor Uri Hasson is a faculty member in the Department of Psychology at Princeton University and is affiliated with the Princeton Neuroscience Institute. His research investigates the neural basis of brain-to-brain human communication, natural language processing, and language acquisition in children within real-world contexts. His lab studies neural responses to natural stimuli, such as audio-visual narratives, and has developed models to better understand how cognition manifests in everyday life. Inspired by advances in deep learning, the Hasson Lab aims to create new theoretical frameworks and computational tools to model neural processes involved in cognition, particularly focusing on natural language processing during open-ended conversations. His work involves analyzing intracranial EEG data collected from epileptic patients engaged in natural conversations, seeking to understand shared computational principles and differences between human brain processing and deep neural network models. His research explores whether deep language models can serve as cognitive models to explain natural language processing in the human brain, with findings suggesting that these models provide a biologically feasible framework for studying language neural mechanisms. Professor Hasson has been recognized for his contributions, including being featured in Princeton Alumni Weekly and receiving the NIH Pioneer Award.

Research topics

• Artificial Intelligence
• Computer Science
• Psychology
• Natural Language Processing
• Linguistics
• Neuroscience
• Cognitive psychology
• Communication
• Cognitive science
• Speech recognition
• Machine Learning
• Sociology
• Mathematics
• Social psychology
• Statistics
• Biology
• Human–computer interaction
• Econometrics
• World Wide Web

Selected publications

• Fast-timescale hippocampal processes bridge between slowly unfurling neocortical states during memory search

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-02-13 · 1 citations

  preprintOpen access

  Prior behavioral work showed that event structure plays a key role in our ability to mentally search through memories of continuous naturalistic experience. We hypothesized that, neurally, this memory search process involves a division of labor between slowly unfurling neocortical states representing event knowledge and fast hippocampal-neocortical communication that supports retrieval of new information at transitions between events. To test this, we tracked slow neural state-patterns in a sample of ten patients undergoing intracranial electroencephalography as they viewed a movie and then searched their memories in a structured naturalistic interview. As patients answered questions ("after X, when does Y happen next?"), state-patterns from movie-viewing were reinstated in neocortex; during memory-search, states unfurled in a forward direction. Moments of state-transition were marked by low-frequency power decreases in cortex and preceded by power decreases in hippocampus that correlated with reinstatement. Connectivity-analysis revealed information-flow from hippocampus to cortex underpinning state-transitions. Together, these results support our hypothesis that fast hippocampal processes bridge between slow neocortical states during memory search.

  PublisherDOI

• Beyond topography: Topographic regularization improves robustness and reshapes representations in convolutional neural networks

  ArXiv.org · 2025-07-31

  preprintOpen accessSenior author

  Topographic convolutional neural networks (TCNNs) are computational models that can simulate aspects of the brain's spatial and functional organization. However, it is unclear whether and how different types of topographic regularization shape robustness, representational structure, and functional organization during end-to-end training. We address this question by comparing TCNNs trained with two local spatial losses applied to a penultimate-layer topographic grid: i) Weight Similarity (WS), whose objective penalizes differences between neighboring units' incoming weight vectors, and ii) Activation Similarity (AS), whose objective penalizes differences between neighboring units' activation patterns over stimuli. We evaluate the trained models on classification accuracy, robustness to weight perturbations and input degradation, the spatial organization of learned representations, and development of category-selective "expert units" in the penultimate layer. Both losses changed inter-unit correlation structure, but in qualitatively different ways. WS produced smooth topographies, with correlated neighborhoods. In contrast, AS produced a bimodal inter-unit correlation structure that lacked spatial smoothness. AS and WS training increased robustness relative to control (non-topographic) models: AS improved robustness to image degradation on CIFAR-10, WS did so on MNIST, and both improved robustness to weight perturbations. WS was also associated with greater input sensitivity at the unit level and stronger functional localization. In addition, as compared to control models, both AS and WS produced differences in orientation tuning, symmetry sensitivity, and eccentricity profiles of units. Together, these results show that local topographic regularization can improve robustness during end-to-end training while systematically reshaping representational structure.

  PublisherOA PDFDOI

• Incremental accumulation of linguistic context in artificial and biological neural networks

  Nature Communications · 2025-01-17 · 16 citations

  articleOpen access

  Large Language Models (LLMs) have shown success in predicting neural signals associated with narrative processing, but their approach to integrating context over large timescales differs fundamentally from that of the human brain. In this study, we show how the brain, unlike LLMs that process large text windows in parallel, integrates short-term and long-term contextual information through an incremental mechanism. Using fMRI data from 219 participants listening to spoken narratives, we first demonstrate that LLMs predict brain activity effectively only when using short contextual windows of up to a few dozen words. Next, we introduce an alternative LLM-based incremental-context model that combines incoming short-term context with an aggregated, dynamically updated summary of prior context. This model significantly enhances the prediction of neural activity in higher-order regions involved in long-timescale processing. Our findings reveal how the brain’s hierarchical temporal processing mechanisms enable the flexible integration of information over time, providing valuable insights for both cognitive neuroscience and AI development. This study reveals how the human brain integrates contextual information differently from Large Language Models. A model that combines short-term and long-term context is introduced, improving predictions of neural activity in higher-order brain regions.

  PublisherOA PDFDOI

• Uncovering a Timescale Hierarchy by Studying the Brain in a Natural Context

  Journal of Neuroscience · 2025-03-19 · 7 citations

  articleOpen access1st authorCorresponding

  As we train multiple generations of students to narrowly design clever, carefully controlled experiments in our confined lab spaces, we may fail to notice, as a field, that we have overlooked fundamental aspects of human cognition. This is a first-person account of how our research and understanding of the neural code were forever transformed when we decided to open the lab's door to the natural world. This journey started with the decision to shift from controlled stimuli to natural dynamic and "messy" stimuli. This transition enabled us to focus on how information is accumulated and processed over time. As a result, we have discovered a new topographic mapping of the hierarchy of cortical processing timescales. I will conclude with a general observation of the paradigm shift occurring in the field as it increasingly emphasizes the study of the neural processes that underlie human behavior in natural, everyday contexts. I am excited to share this journey with you.

  PublisherOA PDFDOI

• The “Podcast” ECoG dataset for modeling neural activity during natural language comprehension

  Scientific Data · 2025-07-03 · 9 citations

  articleOpen accessSenior author

  Naturalistic electrocorticography (ECoG) data are a rare but essential resource for studying the brain's linguistic capabilities. ECoG offers high temporal resolution suitable for investigating processes at multiple temporal timescales and frequency bands. It also provides broad spatial coverage, often along critical language areas. Here, we share a dataset of nine ECoG participants with 1,330 electrodes listening to a 30-minute audio podcast. The richness of this naturalistic stimulus can be used for various research questions, from auditory perception to narrative integration. In addition to the neural data, we extracted linguistic features of the stimulus ranging from phonetic information to large language model word embeddings. We use these linguistic features in encoding models that relate stimulus properties to neural activity. Finally, we provide detailed tutorials for preprocessing raw data, extracting stimulus features, and running encoding analyses that can serve as a pedagogical resource or a springboard for new research.

  PublisherOA PDFDOI

• Music-evoked reactivation during continuous perception is associated with enhanced subsequent recall of naturalistic events

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-07-06

  preprintOpen access

  Music is a potent cue for recalling personal experiences, yet the neural basis of music-evoked memory remains elusive. We address this question by using the full-length film Eternal Sunshine of the Spotless Mind to examine how repeated musical themes reactivate previously encoded events in cortex and shape next-day recall. Participants in an fMRI study viewed either the original film (with repeated musical themes) or a no-music version. By comparing neural activity patterns between these groups, we found that music-evoked reactivation of neural patterns linked to earlier scenes in the default mode network was associated with improved subsequent recall. This relationship was specific to the music condition and persisted when we controlled for a proxy measure of initial encoding strength (spatial intersubject correlation), suggesting that music-evoked reactivation may play a role in making event memories stick that is distinct from what happens at initial encoding.

  PublisherOA PDFDOI

• A Predictive Processing Framework for Joint Action and Communication

  2025-02-01 · 4 citations

  preprintOpen accessSenior author

  Humans act together to achieve feats they could never achieve alone and communicate to ensure alignment of meaning and understanding across different individuals. Explaining the unique human joint action and communication abilities poses an enormous challenge because it requires a systematic account of how people go beyond their own individual perceptions, thoughts, and needs to achieve joint outcomes and align their understanding. Here, we advance a new unified computational framework for explaining joint action and communication. It builds upon influential predictive processing architectures, extending them from individual cognition to multiagent, interactive settings. We assume that joint action and communication involve using and updating agent-neutral models that enable co-agents to predict collective outcomes of interactions regardless of who achieved them. This contrasts with previous frameworks postulating that agent-specific models predict action outcomes for self and others. We discuss three key claims derived from our framework: 1) Co-agents use agent-neutral predictive frameworks during joint action; 2) Co-agents update agent-neutral models interactively by shaping others’ predictions through verbal and non-verbal communication; and 3) Agent-neutral models enable dynamic role allocation during joint action. We highlight how these three claims stem from our proposal, what evidence currently favors or disfavors them, and what novel experiments could be conducted to test them further. Our agent-neutral predictive processing framework will provide a new perspective for understanding the individual basis of human sociality, which closely links theories of joint action and communication to principles of computational neuroscience.

  PublisherDOI

• Temporal structure of natural language processing in the human brain corresponds to layered hierarchy of large language models

  Nature Communications · 2025-11-26 · 3 citations

  articleOpen accessSenior author

  Large Language Models (LLMs) offer a framework for understanding language processing in the human brain. Unlike traditional models, LLMs represent words and context through layered numerical embeddings. Here, we demonstrate that LLMs' layer hierarchy aligns with the temporal dynamics of language comprehension in the brain. Using electrocorticography (ECoG) data from participants listening to a 30-minute narrative, we show that deeper LLM layers correspond to later brain activity, particularly in Broca's area and other language-related regions. We extract contextual embeddings from GPT-2 XL and Llama-2 and use linear models to predict neural responses across time. Our results reveal a strong correlation between model depth and the brain's temporal receptive window during comprehension. We also compare LLM-based predictions with symbolic approaches, highlighting the advantages of deep learning models in capturing brain dynamics. We release our aligned neural and linguistic dataset as a public benchmark to test competing theories of language processing.

  PublisherOA PDFDOI

• A Predictive Processing Framework for Joint Action and Communication

  2025-08-07 · 1 citations

  preprintOpen accessSenior author

  Humans act together to achieve feats they could never achieve alone and communicate to ensure alignment of meaning and understanding across different individuals. Explaining the unique human joint action and communication abilities poses an enormous challenge because it requires a systematic account of how people go beyond their own individual perceptions, thoughts, and needs to achieve joint outcomes and align their understanding. Here, we advance a new unified computational framework for explaining joint action and communication. It builds upon influential predictive processing architectures, extending them from individual cognition to multiagent, interactive settings. We assume that joint action and communication involve using and updating agent-neutral models that enable co-agents to predict collective outcomes of interactions regardless of who achieved them. This contrasts with previous frameworks postulating that agent-specific models predict action outcomes for self and others. We discuss three key claims derived from our framework: 1) Co-agents use agent-neutral predictive frameworks during joint action; 2) Co-agents update agent-neutral models interactively by shaping others’ predictions through verbal and non-verbal communication; and 3) Agent-neutral models enable dynamic role allocation during joint action. We highlight how these three claims stem from our proposal, what evidence currently favors or disfavors them, and what novel experiments could be conducted to test them further. Our agent-neutral predictive processing framework will provide a new perspective for understanding the individual basis of human sociality, which closely links theories of joint action and communication to principles of computational neuroscience.

  PublisherOA PDFDOI

• The “Podcast” ECoG dataset for modeling neural activity during natural language comprehension

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-02-16

  preprintOpen accessSenior author

  Naturalistic electrocorticography (ECoG) data are a rare but essential resource for studying the brain's linguistic capabilities. ECoG offers a high temporal resolution suitable for investigating processes at multiple temporal timescales and frequency bands. It also provides broad spatial coverage, often along critical language areas. Here, we share a dataset of nine ECoG participants with 1,330 electrodes listening to a 30-minute audio podcast. The richness of this naturalistic stimulus can be used for various research endeavors, from auditory perception to semantic integration. In addition to the neural data, we extract linguistic features of the stimulus ranging from phonetic information to large language model word embeddings. We use these linguistic features in encoding models that relate stimulus properties to neural activity. Finally, we provide detailed tutorials for preprocessing raw data, extracting stimulus features, and running encoding analyses that can serve as a pedagogical resource or a springboard for new research.

  PublisherOA PDFDOI

Recent grants

• Neural dynamics supporting integration and recall over long timescales during natural continuous input

  NIH · $528k · 2017–2021

• Brain-to-brain dynamical Coupling: A New framework for the communication of social knowledge

  NIH · $2.0M · 2017–2024

• Speaker-listener coupling: a novel neural approach for assessing communication

  NIH · $4.5M · 2016–2024

• Speaker-listener coupling: a novel neural approach for assessing communication

  NIH · $1.1M · 2016–2021

• Topographic mapping of a hierarchy of temporal receptive windows using natural st

  NIH · $2.0M · 2011–2017

Frequent coauthors

• Kenneth A. Norman

  Princeton University

  98 shared

• Samuel A. Nastase

  Princeton University

  63 shared

• Janice Chen

  Johns Hopkins University

  46 shared

• Rafael Malach

  Weizmann Institute of Science

  44 shared

• Bobbi Aubrey

  Princeton University

  38 shared

• Christopher J. Honey

  Johns Hopkins University

  35 shared

• Zaid Zada

  Princeton University

  32 shared

• Erez Simony

  Holon Institute of Technology

  31 shared

Education

• PhD, Neurobiology

  Weizmann Institute of Science

  2004

Awards & honors

• NIH Pioneer Award