As the head of the Computational Neurobiology Laboratory at Salk, Sejnowski’s latest project is poised to delve deeply into the intricate workings of neural circuits involved in working memory. With the NIH award funding $3.5 million over the next five years, his team intends to revolutionize our understanding by leveraging advanced computational models to dissect the dynamical activity of neuronal networks. These models aim to uncover principles that govern how information is temporarily encoded and maintained in the brain, a cornerstone process critical for decision-making and reasoning.

Working memory, a fundamental cognitive system, enables the transient storage and manipulation of information necessary for complex tasks. Despite its significance, the precise neural mechanisms underlying its robustness and capacity limitations remain enigmatic. Sejnowski’s approach utilizes state-of-the-art techniques in computational neuroscience, combining biophysically realistic neural network models with machine learning algorithms to simulate and predict the behavior of large-scale brain circuits. Such technological sophistication promises to elucidate how synaptic plasticity and circuit dynamics coalesce to sustain memory traces in the prefrontal cortex and hippocampus.

This research direction extends beyond theoretical exploration by aiming to address practical concerns related to neuropsychiatric conditions. Memory impairments are pervasive across multiple disorders, including schizophrenia, traumatic brain injury (TBI), and post-traumatic stress disorder (PTSD). By decoding the neural code of working memory with unprecedented granularity, Sejnowski’s team aspires to identify biomarkers and potential therapeutic targets that could alleviate cognitive deficits associated with these afflictions. The translational impact of this project cannot be overstated as it offers hope for interventions that restore normal cognitive function.

Throughout his distinguished career, Sejnowski has consistently pioneered transformative technologies and ideas in neuroscience. His earlier contributions profoundly shaped the study of neuroeconomics, neuroanatomy, neurophysiology, cognitive psychology, and artificial intelligence. A landmark achievement was his 1985 invention of the Boltzmann machine alongside Geoffrey Hinton, PhD, a quantum leap for neural network learning algorithms. The Boltzmann machine represents the first algorithm capable of learning internal representations in multilayer networks, providing a biologically plausible framework that continues to influence the development of AI to this day.

Beyond algorithmic advances, Sejnowski developed NETtalk, an early neural network system that mimicked human speech synthesis by learning to convert written text into phonetic output. This pioneering work laid a critical foundation for the field of deep learning and current technologies like ChatGPT, which simulate language and cognition through complex neural architectures. His dual focus on biological inspiration and computational implementation epitomizes the fusion of neuroscience and artificial intelligence.

Sejnowski’s influence extends deeply into neuroimaging advancements and the exploration of behaviorally relevant neural systems. His recent innovation involves a novel method for precisely measuring synaptic strength and plasticity, vital parameters underpinning learning and memory formation. By quantifying how synapses encode information and undergo activity-dependent changes, this technique sheds light on the synaptic basis of cognitive decline observed in aging and neurodegenerative diseases. Such insights are critical for developing strategies to preserve cognitive function across the lifespan.

Recognition of Sejnowski’s scientific excellence is reflected in the numerous prestigious awards he has garnered over the years. Most notably, he was honored with the 2024 Brain Prize and the 2022 Gruber Prize in Neuroscience, reflecting his outstanding contributions to understanding brain function. These accolades underscore his role as a luminary whose work bridges multiple disciplines and drives progress in the neurosciences.

A testament to his standing in the scientific community, Sejnowski holds positions in several elite academies, including the United States National Academy of Sciences, National Academy of Medicine, National Academy of Engineering, and National Academy of Inventors. He is also a Fellow of the Royal Society in the United Kingdom and a member of the American Philosophical Society, highlighting his global influence as a thought leader and innovator.

The Salk Institute, where Sejnowski conducts his research, is renowned for its relentless pursuit of groundbreaking knowledge in biological sciences. Founded by Jonas Salk, the developer of the first effective polio vaccine, the institute embodies a culture of bold scientific exploration. Within its walls, scientists continually push the boundaries of understanding in neuroscience, cancer biology, aging, immunobiology, and computational biology, among other fields. Sejnowski’s work fits seamlessly into this mission, combining innovation with impact.

Sejnowski’s ongoing commitment to deciphering the biological basis of cognition, learning, and memory signifies a new chapter in brain science. As computational tools grow increasingly sophisticated, melding detailed biophysical models with AI-driven analytics, his laboratory stands at the forefront of unraveling the complexities of neural information processing. This work holds profound implications not only for fundamental neuroscience but also for the development of targeted treatments for cognitive impairments.

With the NIH Director’s Pioneer Award supporting this ambitious endeavor, Terrence Sejnowski’s visionary research will continue to illuminate the neural architecture of working memory. His innovative approach exemplifies the power of interdisciplinary science to tackle some of the most challenging questions in brain function and dysfunction. The coming years promise remarkable advances from his lab that could redefine therapeutic strategies for mental health disorders and enhance our comprehension of the human mind.

Subject of Research: Neural circuit dynamics of working memory and computational modeling of cognitive processes in health and disease.

Article Title: NIH Director’s Pioneer Award Enables Terrence Sejnowski’s Cutting-Edge Computational Exploration of Working Memory Circuitry

News Publication Date: October 13, 2025

Web References:
– https://www.salk.edu/scientist/terrence-sejnowski/
– https://commonfund.nih.gov/pioneer
– https://www.salk.edu/news-release/upgrading-brain-storage-quantifying-how-much-information-our-synapses-can-hold/
– https://www.salk.edu/news-release/salk-professor-terrence-sejnowski-wins-brain-prize/
– https://www.salk.edu/news-release/salk-institutes-terrence-sejnowski-awarded-gruber-prize/
– https://www.salk.edu/news-release/terrence-sejnowski-elected-to-the-royal-society-and-the-american-philosophical-society/

Image Credits: Salk Institute

Keywords: Life sciences, Neuroscience, Cognitive neuroscience, Computational neuroscience, Memory, Cognition, Cognitive psychology, Psychological science, Social sciences

Traditional reinforcement learning (RL) models have long simplified reward prediction by collapsing the expectation of future outcomes into a single average value. While this approach provides a usable heuristic, it glosses over the intricate variability present in real-world scenarios, where timing and reward size fluctuate unpredictably. The dopamine system—a network of neurons releasing the neurotransmitter dopamine—has been central to this framework, signaling when outcomes exceed or fall short of expectations. Yet, these models have struggled to account for the nuanced ways organisms navigate uncertainty, impulsivity, and risk-sensitive behaviors.

The compelling innovation presented by the Champalimaud team lies in uncovering how diverse groups of dopamine neurons collectively map a richer distribution of potential rewards. By integrating both magnitude and temporal dimensions of reward prediction, these neurons form a neural coordinate system capable of encoding not just if, but when and how much reward might be delivered. Such a multidimensional representation allows for far more flexible and context-sensitive decision-making than previous models suggest, providing a biological foundation for nuanced behavioral adaptations.

This research draws inspiration from and intersects with contemporary AI methodologies, particularly those involving distributional reinforcement learning. Unlike classical RL that computes a mean expected reward, distributional RL algorithms generate full probability distributions over possible outcomes, enabling machines to manage uncertainty and risk more adeptly. The Champalimaud study thus closes a loop by showing that the brain’s dopamine circuitry might be implementing a natural analog of these AI strategies, suggesting a profound evolutionary convergence on efficient learning principles.

Experimental evidence came from carefully designed tasks involving mice exposed to olfactory cues predicting rewards that varied not only in size but in timing. Instead of averaging neuronal responses, the researchers leveraged sophisticated genetic labeling and decoding techniques to observe the diversity among individual dopamine neurons. Their findings were striking: certain neurons exhibited “impatient” profiles, emphasizing immediate rewards, while others were more attuned to delayed gratification. Similarly, some neurons demonstrated an “optimistic” bias towards unexpectedly large rewards, while others adopted a “pessimistic” stance, favoring cautious estimates to minimize risk.

Collectively, these differentiated tuning properties among dopamine neurons construct a full probabilistic landscape that animals can reference when making decisions. The map is dynamic, showing impressive adaptability—neurons recalibrate their sensitivity based on environmental contexts. For instance, when rewards are typically delayed, the neuronal population shifts its coding to elevate the importance of future, later-arriving rewards. This flexible “efficient coding” mechanism mirrors principles seen in sensory and cognitive systems, optimizing resource use and behavioral output in fluctuating conditions.

One of the most fascinating implications of this neural architecture is its conceptual analogy to a team of advisors, each with unique risk preferences. This ensemble approach is widespread in AI, where models operating with different biases or viewpoints collaborate to improve prediction accuracy under uncertainty. In the brain, such neuronal diversity appears critical to navigating the unpredictable complexities of the environment, balancing the urge to act swiftly against the wisdom of patience.

Beyond explaining typical decision-making, the study casts new light on impulsivity and self-control. Variability in the dopamine system’s representation of future rewards might underpin why some individuals are more prone to immediate gratification, while others consistently defer reward for potentially greater gains. This insight opens avenues for targeted interventions that could “reshape” this neural map through behavioral therapies or environmental manipulation, fostering healthier risk evaluation and impulse regulation.

Computational simulations complementing the biological work underscored the functional utility of this multidimensional dopamine code. Artificial agents equipped with access to these dopamine-like maps outperformed traditional models, particularly in variable and shifting environments. Notably, the ability to reweight the importance of different future outcomes rapidly—without constructing exhaustive world models—affords elegant and efficient adaptation, crucial for survival and success in complex real-world settings.

Intriguingly, the study highlights that such rich dopaminergic coding occurs early, at the moment of cue presentation, before any reward delivery. This timing suggests that the brain does not merely react to past outcomes but proactively anticipates a distribution of possible futures, enabling more strategic planning and foresight. Understanding this temporal structure in dopamine neuron activity offers profound insights into the neural basis of predictive cognition and goal-directed behavior.

The synergy between neuroscience and AI embodied in this research signals a new horizon for both fields. By elucidating how biological systems naturally encode full distributions of possible futures, we glean not only fundamental knowledge about brain function but also inspiration for engineering smarter AI. Machines designers increasingly aim to replicate these multifaceted predictive capabilities to allow artificial agents to navigate uncertainty, adapt to shifting goals, and make decisions that resemble human judgment more closely.

At its core, the discovery of a dopamine-coded probabilistic map positions the brain as a masterful architect of foresight, weaving together complexity, flexibility, and diversity into a neural framework that empowers organisms to thrive amid uncertainty. Rather than a fixed forecast, the future emerges as a landscape of possibilities dynamically shaped by experience and context. This work not only enriches our scientific understanding but also kindles optimism for innovations in mental health, AI development, and beyond—a vivid reminder of how fundamental research can illuminate the intricate dance between biology and technology.

For future inquiries, the implications are profound and manifold. Advancing this line of research may unravel the neural substrates of various neuropsychiatric conditions marked by dysfunctional reward processing, such as addiction or compulsive behaviors. Moreover, AI systems inspired by these biological principles might soon engender new generations of adaptive, risk-aware technologies that extend into domains as diverse as autonomous vehicles, personalized medicine, and complex strategic planning.

As you next ponder whether to wait in line for your favorite meal or settle for an available snack, remember this: hidden in your brain is a sophisticated, multidimensional map crafted by dopamine neurons. It charts not only the reward that awaits but the myriad ways that reward might come to you—timing, size, and probability—all converging to steer your choice. This neural compass, reflecting millions of years of evolution and recently mirrored in the frontiers of AI, continues to guide each moment of decision-making in the ever-uncertain journey of life.

Subject of Research: Animals
Article Title: Dopamine neurons encode a multidimensional probabilistic map of future reward
News Publication Date: 4-Jun-2025
Web References: 10.1038/s41586-025-09089-6
Image Credits: Joe Paton
Keywords: Dopamine, Dopaminergic neurons, Neurotransmitters, Artificial intelligence, Machine learning, Artificial neural networks, Learning, Brain, Neural modeling, Computational neuroscience, Decision making, Risk perception, Probability distributions

The central serotonin (5-HT) system has long been recognized as critical to a host of neurological and psychiatric functions, including mood regulation, anxiety, and reward processing. Prevailing models have treated individual serotonin neurons as largely autonomous units, functioning independently within the raphe nuclei of the midbrain. However, the University of Ottawa-led team has directly demonstrated that these neurons are instead interconnected, forming complex recurrent networks through axonal projections. This newly identified synaptic connectivity creates a dynamic feedback architecture, facilitating nonlinear inhibitory mechanisms that finely tune serotonin release across diverse brain regions.

Leveraging an innovative combination of electrophysiology, cellular imaging, optogenetic manipulation, and behavioral assays, the research team dissected the functional organization of serotonin neurons at an unprecedented resolution. Complementing these experimental approaches, advanced mathematical modeling and computational simulations provided critical insights into the emergent properties of these networks. The data indicate that distinct ensembles of serotonin neurons exhibit unique temporal activity patterns, allowing for regionally specific modulation of serotonin release that deviates markedly from a uniform broadcast signal.

One of the most striking implications of this work is the revision it forces upon the “winner-takes-all” neural computation framework previously applied to serotonergic circuits. Instead of a simple competitive selection among neurons, the identified recurrent inhibition mechanism suggests that highly active serotonin ensembles can suppress the output of less active groups. This dynamic antagonism introduces a sophisticated layer of regulation, enabling the brain to perform nuanced, context-dependent decision computations related to risk, threat assessment, and behavioral choice.

The lateral habenula, a small but powerful brain region engaged during aversive experiences, emerges as a key modulator of this intricate serotonin network. Known to encode environmental threat signals and implicated in the pathophysiology of major depressive disorder, the habenula’s influence on raphe serotonin neurons underscores a neurobiological substrate for how perceived dangers shape binary decision-making processes. Through this circuitry, the brain evaluates scenarios such as whether to proceed with a risky action or avoid a dangerous environment, fundamentally guiding everyday behavioral choices from utter avoidance to bold engagement.

Dr. Jean-Claude Béïque, senior investigator and professor at the University of Ottawa, emphasizes how this holistic interpretation of serotonergic function reshapes therapeutic perspectives. “Our findings dismantle the outdated notion of serotonin neurons acting as isolated messengers," he states. "Acknowledging the intricate, recurrent interplay among these neurons opens novel avenues for targeted interventions in mood disorders, potentially refining treatments for conditions like depression by focusing on circuit-level dynamics rather than diffuse neurotransmitter modulation.”

The study’s first author, Dr. Michael Lynn, who completed his doctoral training at the University of Ottawa and is now conducting postdoctoral research at the University of Oxford, highlights the methodological rigor underpinning these discoveries. “By integrating optogenetics with real-time calcium imaging and electrophysiological recordings, we captured the temporal sequencing of serotonin neuron activity in living organisms navigating decision paradigms,” he explains. “Our behavioral analyses, although initially conducted in controlled experimental conditions, suggest these complex inhibitory interactions facilitate adaptive, flexible choices amidst ambiguous or conflicting sensory inputs.”

Mathematical modeling played a pivotal role in deciphering the nonlinear dynamics underlying serotonin release patterns. The team employed dynamical systems theory and network modeling to simulate how recurrent inhibitory loops produce emergent phenomena such as activity-dependent suppression and facilitation. These computational insights matched experimental observations, validating the hypothesis that serotonin neurons operate within an intricate feedback system — rather than as isolated transmitters — to implement circuit-level decision computations.

Furthermore, the discovery sheds light on the previously enigmatic heterogeneity of the serotonin system. Instead of a monolithic neurotransmitter network broadcasting a uniform modulatory tone, the identified subpopulations of serotonin neurons form partially independent ensembles tuned to distinct brain targets. This spatial and temporal differentiation in serotonergic signaling likely enables the brain to finely balance multiple motivational, emotional, and cognitive demands simultaneously, expanding the functional repertoire of the central serotonin system beyond conventional conceptualizations.

Looking forward, the Ottawa research team is poised to extend these findings through behavioral studies in naturalistic settings. Their goal is to determine whether the nonlinear recurrent inhibition mechanisms identified in simplified experimental contexts also govern serotonin-mediated decision making during complex, ecologically valid behaviors in rodents. This translational approach holds promise for connecting cellular and circuit-level discoveries to whole-organism function, with profound implications for understanding psychiatric disease states rooted in serotonergic dysregulation.

This paradigm-shifting work not only elevates serotonin research into a new era of circuit neuroscience but also bridges experimental neuroscience with sophisticated computational frameworks. By deciphering how neuronal ensembles compute binary decisions through recurrent inhibition and facilitation, the study unveils fundamental principles of neural information processing that transcend serotonin signaling alone. These insights could inspire novel algorithmic strategies in artificial intelligence, where biologically inspired neural networks emulate the competitive and cooperative dynamics exhibited by serotonin ensembles.

In sum, the University of Ottawa team’s multidisciplinary approach has carved a transformative path in neuroscience, revealing that serotonin neurons in the raphe nuclei function as interlinked networks employing nonlinear feedback to orchestrate decision-making choices at the neural circuit level. This reconceptualization invites a reassessment of how neuromodulators shape cognition and behavior and promises to impact diverse fields from clinical psychiatry to computational neuroscience and beyond.

Subject of Research: Cells

Article Title: Nonlinear recurrent inhibition through facilitating serotonin release in the raphe

News Publication Date: 2-Apr-2025

Web References:

• University of Ottawa Faculty of Medicine https://www.uottawa.ca/faculty-medicine
• Dr. Jean-Claude Béïque Profile https://www.uottawa.ca/faculty-medicine/dr-jean-claude-beique
• Nature Neuroscience Article DOI: 10.1038/s41593-025-01912-7

Image Credits: Faculty of Medicine, University of Ottawa

Keywords: Serotonin, Computational neuroscience, Serotonin receptor signaling, Social decision making, Molecular neuroscience, Adenylate cyclase activity, Cellular processes, Mathematical modeling, Light, Midbrain, Cognitive function, Network modeling, Neural modeling, Signaling complexes, Dynamical systems, Neurotransmitters