Biology subjects

Hiratani, N.

Publications and source records attributed to Hiratani, N..

10 recordsLinked to original sources

bioRxiv2026

Local gated-Hebbian learning of deep cerebellar networks with quadratic classification capacity

A central goal of neuroscience is to understand how neural circuit architecture supports learning. While recent work has clarified the computational role of depth in sensory cortical hierarchies, it remains unclear why predominantly feedforward, non-convolutional circuits such as the cerebellum and olfactory system also contain multiple processing layers. Theoretical work in deep learning has shown that two-hidden-layer networks can achieve classification capacity that scales quadratically with the number of intermediate neurons, but these results rely on nonlocal synaptic optimization and are therefore difficult to reconcile with biological learning rules. Here, we show analytically and numerically that a two-hidden-layer network with feedforward gating can achieve quadratic capacity using local three-factor Hebbian learning when intermediate activity is sparse. This architecture supports efficient one-shot learning and, in settings where backpropagation requires many repeated weight updates, offers an advantage in learning speed. Beyond random perceptron tasks, the model also performs well on structured cerebellum-related tasks, including reinforcement-learning-based motor control. Mapping the model onto cerebellar microcircuitry further suggests functional roles for dendritic compartmentalization, branch-specific inhibition, and disinhibitory interneuron pathways. Together, these results extend the Marr-Albus-Ito framework by showing how the presence of multiple intermediate layers in cerebellum-like circuits can support fast, local, and high-capacity learning.

neuroscience↗

bioRxiv2026

Sex differences in exploration-exploitation strategies during home-cage decision making

The exploration-exploitation trade-off refers to the conflict between using known strategies that reliably yield reward (exploitation) and sampling uncertain options that might yield better outcomes (exploration). Dysregulation of this balance is implicated in neuropsychiatric disease, and while sex differences in this balance have been described, the biological bases remain unclear. To quantify sex differences in this trade-off, we tested mice (n=74 male, 62 female) on four home-cage based foraging tasks with an operant pellet dispensing device, Feeding Experimentation Device 3 (FED3). Mice completed the tasks continuously over multiple days and the tasks were their only source of food. Across multiple tasks, males showed higher win-stay behaviour than females, indicating greater exploitation of previously rewarded actions, an effect that was modest in size but highly significant. Power analyses revealed that >30 mice per sex were needed to detect these modest but significant sex differences with 80% power. No consistent sex differences were observed in pellet intake, suggesting that differences in exploitation did not reflect differences in hunger drive or demand for pellets. Exploitation is a more efficient strategy when environmental parameters are fixed, while exploration can be more advantageous when parameters such as reward locations are changing and uncertain. We tested this idea by re-running our mice in a probabilistic foraging task, where actions led to uncertain probabilities of reward. While males continued to show higher levels of win-stay behaviour on this task, this no longer led to increases in accuracy. Behavioural modelling also supported this framework, demonstrating that stronger win-stay behaviour was most advantageous in deterministic models, and less advantageous in probabilistic models. Together, our findings demonstrate that male and female mice have small but significant differences in their exploration-exploitation balance, which leads to more accurate foraging in certain, but not uncertain, environments.

neuroscience↗

bioRxiv2025

Ventral striatal astrocytes contribute to reinforcement learning

Astrocytes influence synaptic plasticity and neuronal function through astrocytic calcium dynamics (ACD). However, astrocytes contribution to cognitive operations like reinforcement learning (RL) remains unclear. To examine this, we trained mice on a RL dependent probabilistic decision-making task. We attenuated ACD across distinct striatal regions, and found ACD attenuation specifically in ventral striatum (VS) increased decision noisiness and impaired reward-guided choice performance. This effect was largely due to a reduction in "win-stay" behavior. Using in-vivo calcium imaging, we found that VS ACD correlated with reward prediction errors (RPEs). Furthermore, these trial-by-trial ACD fluctuations predicted trial-by-trial choice variability. In-silico lesions of a biologically constrained circuit model suggest that astrocytes could regulate behavioral variability by sharing RPE signals across populations of striatal neurons. Together, these results suggest that VS astrocytes contribute to cortico-striatal functions to mediate decision noisiness.

neuroscience↗

bioRxiv2025

Cortical neural landscape captures mouse-to-mouse variability in anticipatory vs. inattentive decision making

Understanding individual variability in behavior is crucial for both basic and clinical neuroscience, yet it remains challenging to study in traditional single laboratory experiments with small sample size. Leveraging standardized behavioral and neural datasets from the International Brain Laboratory, comprising approximately 100 mice trained on a visual decision-making task, we investigated the structure and neural correlates of inter-animal behavioral variability. Using reaction time analysis and a deep learning-based embedding of individual animals, we uncovered large but low-dimensional differences in behavioral traits. Some mice consistently exhibited anticipatory responses, marked by fast reaction times, while others showed slower, more disengaged behavior. These behavioral profiles were consistent across sessions, with female mice tending to show more anticipatory behavior than males. We hypothesized that this behavioral spectrum reflects differences in the depth of underlying cortical states, reflected in the temporal dynamics of neural activity. Supporting this idea, we found that the characteristic timescale of population activity, measured during both inter-trial intervals and passive periods, correlated with an animals anticipatory tendency across cortical areas, especially in medial visual areas. These findings suggest that individual differences in the cortical dynamics may underlie distinct decision-making strategies.

neuroscience↗

bioRxiv2023

Time Tiling of Learned Stimulus-Based Decision Making in Dorsal CA1 in the Hippocampus

Sequential neural dynamics encoded by "time cells" play a crucial role in hippocampal function. However, the role of hippocampal sequential neural dynamics in associative learning is an open question. In this manuscript, we used two-photon Ca2+ imaging of dorsal CA1 pyramidal neurons in head-fixed mice performing a go-no-go associative learning task. We found that pyramidal cells responded differentially to the rewarded or unrewarded stimuli. The stimuli were decoded accurately from the activity of the neuronal ensemble, and accuracy increased substantially as the animal learned to differentiate the stimuli. Decoding the stimulus from individual pyramidal cells that responded differentially revealed that decision-making took place at discrete times after stimulus presentation. Lick prediction decoded from the ensemble activity of cells in dCA1 correlated linearly with lick behavior indicating that sequential activity of pyramidal cells in dCA1 constitutes a temporal memory map used for decision-making in associative learning.

neuroscience↗

bioRxiv2022

Evolution of neural activity in circuits bridging sensory and abstract knowledge

The ability to associate sensory stimuli with abstract classes is critical for survival. How are these associations implemented in brain circuits? And what governs how neural activity evolves during abstract knowledge acquisition? To investigate these questions, we consider a circuit model that learns to map sensory input to abstract classes via gradient descent synaptic plasticity. We focus on typical neuroscience tasks (simple, and context-dependent, categorization), and study how both synaptic connectivity and neural activity evolve during learning. To make contact with the current generation of experiments, we analyze activity via standard measures such as selectivity, correlations, and tuning symmetry. We find that the model is able to recapitulate experimental observations, including seemingly disparate ones. We determine how, in the model, the behaviour of these measures depends on details of the circuit and the task. These dependencies make experimentally-testable predictions about the circuitry supporting abstract knowledge acquisition in the brain.

neuroscience↗

bioRxiv2020

Developmental and evolutionary constraints on olfactory circuit selection

Across species, neural circuits show remarkable regularity, suggesting that their structure has been driven by underlying optimality principles. Here, we ask whether we can predict the neural circuitry of diverse species by optimizing the neural architecture to make learning as efficient as possible. We focus on the olfactory system, primarily because it has a relatively simple evolutionarily conserved structure, and because its input and intermediate layer sizes exhibits a tight allometric scaling. In mammals, it has been shown that the number of neurons in layer 2 of piriform cortex scales as the number of glomeruli (the input units) to the 3/2 power; in invertebrates, we show that the number of mushroom body Kenyon cells scales as the number of glomeruli to the 7/2 power. To understand these scaling laws, we model the olfactory system as a three layered nonlinear neural network, and analytically optimize the intermediate layer size for efficient learning from a limited number of samples. We find that the 3/2 scaling observed in mammals emerges naturally, both in full batch optimization and under stochastic gradient learning. We extended the framework to the case where a fraction of the olfactory circuit is genetically specified, not learned. We show numerically that this makes the scaling law steeper when the number of glomeruli is small, and we are able to recover the 7/2 scaling law observed in invertebrates. This study paves the way for a deeper understanding of the organization of brain circuits from an evolutionary perspective.

neuroscience↗

bioRxiv2019

Rapid Bayesian learning in the mammalian olfactory system

Many experimental studies suggest that animals can rapidly learn to identify odors and predict the rewards associated with them. However, the underlying plasticity mechanism remains elusive. In particular, it is not clear how olfactory circuits achieve rapid, data efficient learning with local synaptic plasticity. Here, we formulate olfactory learning as a Bayesian optimization process, then map the learning rules into a computational model of the mammalian olfactory circuit. The model is capable of odor identification from a small number of observations, while reproducing cellular plasticity commonly observed during development. We extend the framework to reward-based learning, and show that the circuit is able to rapidly learn odor-reward association with a plausible neural architecture. These results deepen our theoretical understanding of unsupervised learning in the mammalian brain.

neuroscience↗

bioRxiv2017

Chunking sequence information by mutually predicting recurrent neural networks

Interpretation and execution of complex sequences is crucial for various cognitive tasks such as language processing and motor control. The brain solves this problem arguably by dividing a sequence into discrete chunks of contiguous items. While chunking has been accounted for by predictive uncertainty, alternative mechanisms have also been suggested, and the mechanism underlying chunking is poorly understood. Here, we propose a class of unsupervised neural networks for learning and identifying repeated patterns in sequence input with various degrees of complexity. In this model, a pair of reservoir computing modules, each of which comprises a recurrent neural network and readout units, supervise each other to consistently predict others responses to frequently recurring segments. Interestingly, this system generates neural responses similar to those formed in the basal ganglia during habit formation. Our model extends reservoir computing to higher cognitive function and demonstrates its resemblance to sequence processing by cortico-basal ganglia loops.

neuroscience↗

bioRxiv2017

Redundancy in synaptic connections enables neurons to learn optimally

Recent experimental studies suggest that, in cortical microcircuits of the mammalian brain, the majority of neuron-to-neuron connections are realized by multiple synapses. However, it is not known whether such redundant synaptic connections provide any functional benefit. Here, we show that redundant synaptic connections enable near-optimal learning in cooperation with synaptic rewiring. By constructing a simple dendritic neuron model, we demonstrate that with multisynaptic connections, synaptic plasticity approximates a sample-based Bayesian filtering algorithm known as particle filtering, and wiring plasticity implements its resampling process. The derived synaptic plasticity rule accounts for many experimental observations, including the dendritic position dependence of spike-timing-dependent plasticity. The proposed framework is applicable to detailed single neuron models, and also to recurrent circuit models. Our study provides a novel conceptual framework for synaptic plasticity and rewiring.

neuroscience↗