Biology subjects

Goldman, M. S.

Publications and source records attributed to Goldman, M. S..

14 recordsLinked to original sources

bioRxiv2026

Constraints for spatially and temporally precise learning in a neural circuit model of reinforcement learning

Reinforcement learning is a key means by which animals learn appropriate actions in a given context. A large body of work suggests that such learning depends on interactions between cortico-basal ganglia circuits and the midbrain dopaminergic system, yet the underlying circuit mechanisms and plasticity rules are not fully understood. Here we present a biologically plausible, multi-region neural circuit model of songbird vocal learning and map it onto the actor-critic framework of reinforcement learning. In this model, stochastic spiking activity in the cortico-basal ganglia pathway implements action selection and drives behavioral exploration, while the pathways driving midbrain dopaminergic signaling evaluate behavioral outcomes and support a reward prediction error based learning rule that approximates stochastic gradient ascent. The model achieves millisecond-scale precise learning that matches observed behavior. We further use the model to examine two fundamental constraints on biological reinforcement learning. First, dopaminergic reinforcement signals are temporally imprecise, which can cause interference between neurons controlling actions that occur close in time. Second, dopaminergic signals are spatially imprecise, which can cause interference between neurons controlling different aspects of behavior but receiving a common reinforcement signal. By jointly modeling the actor and critic components of the circuit, we show that fast updating of reward prediction is crucial for precise and efficient learning under both forms of interference, and the model predicts the experimentally observed timescale of reward prediction updating. These results suggest a circuit-level mechanism by which biological systems achieve reinforcement learning despite the temporal and spatial limitations of global neuromodulatory signals.

neuroscience↗

bioRxiv2025

Transfer of graded information through gated receptivity to widely broadcast signals

Making accurate decisions requires the brain to maintain evolving representations of the evidence supporting different possible choices. The population of neurons that maintains this evolving representation may change over the course of the decision-making process. Such changes may result from a switch in the set of neurons representing the relevant behavioral output or a switch in the set of neurons receiving task-relevant information. Recent work has shown how intervening actions like eye movements or navigation to a new location can shift the set of neurons that encode subsequent inputs and outputs. Here, we present a computational model of how analog-valued information that supports an evolving decision can be flexibly transferred between populations of neurons without changing the synaptic connectivity of the underlying network. This is accomplished through a dynamic gating mechanism in which information is widely broadcast throughout a network, but only those neurons receiving a gating signal are receptive to this information. Using this framework, we provide a mechanistic explanation for recent experimental results in which information supporting an evolving decision was shown to be transferred between different populations of parietal cortex neurons during intervening smooth pursuit and saccadic eye movements. This mechanism enables organisms to maintain a continuous decision process despite changing frames of reference, offering a potentially general framework for cognitive continuity in dynamic environments.

neuroscience↗

bioRxiv2025

Robust maintenance of both stimulus location and amplitude in a working memory model based on dendritic bistability

Working memory is a core feature of cognition that enables items to be maintained and manipulated over short durations of time. Stored information can be binary, such as the presence or absence of an object, or graded, such as the graded intensity or location of a feature. Current computational models of working memory cannot robustly maintain both the graded intensity and spatial location of a stored item. Here, we show how this limitation can be overcome if neurons contain multiple bistable dendritic compartments. First, we illustrate the core mechanism for the storage of graded amplitude information in a simple spiking "autapse" circuit consisting of a single neuron connected to itself. Second, we reduce this model to a rate-based model that permits analytic understanding. Third, we implement this mechanism within a spatially extended architecture in which the spatial location of an item is encoded by the set of active neurons. In contrast to classic spatial working memory models, which only encode the binary presence of an item at a given location, the multi-dendrite-neuron model robustly encodes both the amplitude and location of an item in working memory in a noise-resistant manner and without requiring fine tuning of parameters. We show analytically that the key mechanism permitting the storage of amplitude information is equivalent to that of the simpler autapse circuit. This work provides a solution to the problem of encoding graded information in spatial working memory and demonstrates how dendritic computation can increase the representational capacity and robustness of working memory. Significance StatementAnimals can readily remember both the location of an item and analog features of the item such as its amplitude or intensity. Remarkably, current computational models of working memory require extreme fine tuning of model parameters to perform this task. Here, we show how this limitation can be overcome if neurons have active dendritic processes that enable local dendritic compartments of the neuron to robustly maintain digital "up" ("plateau potential") or "down" states. By activating a variable number of plateau potentials at each location, the models can jointly maintain both the location and amplitude of a stimulus in working memory in a noise-resistant manner. This work demonstrates how local dendritic processes can enhance the computational capabilities of working memory networks.

neuroscience↗

bioRxiv2024

Context-dependence of deterministic and nondeterministic contributions to closed-loop steering control

In natural circumstances, sensory systems operate in a closed loop with motor output, whereby actions shape subsequent sensory experiences. A prime example of this is the sensorimotor processing required to align ones direction of travel, or heading, with ones goal, a behavior we refer to as steering. In steering, motor outputs work to eliminate errors between the direction of heading and the goal, modifying subsequent errors in the process. The closed-loop nature of the behavior makes it challenging to determine how deterministic and nondeterministic processes contribute to behavior. We overcome this by applying a nonparametric, linear kernel-based analysis to behavioral data of monkeys steering through a virtual environment in two experimental contexts. In a given context, the results were consistent with previous work that described the transformation as a second-order linear system. Classically, the parameters of such second-order models are associated with physical properties of the limb such as viscosity and stiffness that are commonly assumed to be approximately constant. By contrast, we found that the fit kernels differed strongly across tasks in these and other parameters, suggesting context-dependent changes in neural and biomechanical processes. We additionally fit residuals to a simple noise model and found that the form of the noise was highly conserved across both contexts and animals. Strikingly, the fitted noise also closely matched that found previously in a human steering task. Altogether, this work presents a kernel-based analysis that characterizes the context-dependence of deterministic and non-deterministic components of a closed-loop sensorimotor task. New and noteworthyWe use nonparametric systems identification techniques to assess the context-dependence of deterministic and nondeterministic contributions to a closed-loop behavior. Classical approaches assume a fixed transformation between sensory input and motor output. Here, we reveal strong changes to the measured sensorimotor transformations with behavioral context. In contrast, noise within the transformation exhibited a consistent form across contexts, subjects, and species. Together, this work demonstrates how context affects the systematic and stochastic components of a closed-loop behavior.

neuroscience↗

bioRxiv2024

Synaptic weight dynamics underlying systems consolidation of a memory

Systems consolidation is a common feature of learning and memory systems, in which a long-term memory initially stored in one brain region becomes persistently stored in another region. We studied the dynamics of systems consolidation in simple circuit architectures with two sites of plasticity, one in an early-learning and one in a late-learning brain area. We show that the synaptic dynamics of the circuit during consolidation of an analog memory can be understood as a temporal integration process, by which transient changes in activity driven by plasticity in the early-learning area are accumulated into persistent synaptic changes at the late-learning site. This simple principle naturally leads to a speed-accuracy tradeoff in systems consolidation and provides insight into how the circuit mitigates the stability-plasticity dilemma of storing new memories while preserving core features of older ones. Furthermore, it imposes two constraints on the circuit. First, the plasticity rule at the late-learning site must stably support a continuum of possible outputs for a given input. We show that this is readily achieved by heterosynaptic but not standard Hebbian rules. Second, to turn off the consolidation process and prevent erroneous changes at the late-learning site, neural activity in the early-learning area must be reset to its baseline activity. We propose two biologically plausible implementations for this reset that suggest novel roles for core elements of the cerebellar circuit. Significance StatementHow are memories transformed over time? We propose a simple organizing principle for how long term memories are moved from an initial to a final site of storage. We show that successful transfer occurs when the late site of memory storage is endowed with synaptic plasticity rules that stably accumulate changes in activity occurring at the early site of memory storage. We instantiate this principle in a simple computational model that is representative of brain circuits underlying a variety of behaviors. The model suggests how a neural circuit can store new memories while preserving core features of older ones, and suggests novel roles for core elements of the cerebellar circuit.

neuroscience↗

bioRxiv2023

Blue spectral quality contributes to yield and ascorbic acid content in Micro Tom tomato under sole-source lighting

There is an intricate relationship between the spectral composition of light, plant photosynthetic performance and biomass accumulation. The interaction between plants and the ambient light environment is not only crop-specific but also crucial for maximized yield and nutritional quality. The emergence of LED technology introduced a unique avenue for manipulating the light spectrum, offering new and continued possibilities for optimizing growth conditions. With a focus on tomato cultivation under sole-source lighting, we have investigated the impact of blue spectral quality (waveband composition) on growth and nutritional value of dwarf tomato. Notably, our study revealed that distinct wavebands of blue light (400 nm, 420 nm and 450 nm) can influence the net photosynthetic rate despite unaltered photochemical efficiency. Our research uncovered a correlation whereby shorter wavelengths of blue light increased leaf area, while longer blue wavelengths contributed to greater harvest indices. In addition, we identified a specific blue peak wavelength, 450 nm, that significantly affected chlorophyll composition in leaves and ascorbate levels in fruits. Through these findings, we call attention to the notion that blue spectral quality has a role in shaping both yield and nutritional attributes of dwarf tomato, such as Micro Tom, under sole-source lighting. Overall, our research provides valuable insight into the nuanced interplay between light spectrum, plant physiology, and horticultural outcome.

plant biology↗

bioRxiv2023

Neural circuit models for evidence accumulation through choice-selective sequences

Decision making is traditionally thought to be mediated by neurons that accumulate evidence through persistent activity. However, recent decision-making experiments in rodents have observed neurons across the brain that fire sequentially, rather than persistently, with the subset of neurons in the sequence depending on the animals choice. We developed two new candidate circuit models in which neurons are active sequentially and transfer evidence faithfully to the next active population. One model encodes evidence in the relative firing of two competing chains of neurons, and the other in the network location of a stereotyped pattern ("bump") of neural activity. Neural recordings from four brain regions during an evidence accumulation task revealed that different regions displayed evidence tuning consistent with different candidate models. This work provides a mechanistic explanation for how graded information may be precisely accumulated within and transferred between neural populations, and suggests that different brain regions may accumulate evidence through different circuit mechanisms.

neuroscience↗

bioRxiv2022

Experience Adaptively Tunes the Timing Rules for Associative Plasticity

The brain learns about the statistical relationships between events in the world through associative synaptic plasticity, controlled by the timing between neural events. Here, we show that experience can dramatically alter the timing rules governing associative plasticity and learning. In normally reared mice, the timing requirements for short- and long-term associative plasticity at synapses in the oculomotor cerebellum are precisely matched to the 120 ms delay for visual feedback to the circuit about behavioral errors. This specialization of the plasticity rules for the constraints of a particular circuit and learning task is acquired through experience. In dark-reared mice that never experienced visual feedback about oculomotor errors, synapses defaulted to a coincidence-based plasticity rule, with a corresponding delay in the timing of learned eye movements. This temporal metaplasticity persists into adulthood; when mice reared normally from birth were moved to dark housing as adults, the task-specific timing requirements for plasticity and the temporal accuracy of learning were lost and then re-established when visual experience was restored. Computational modeling suggests two general classes of biologically plausible mechanisms, each with multiple possible implementations, that can tune plasticity to distinct features of the statistics of neural activity. Temporal metaplasticity provides a potentially general mechanism for experience-dependent improvement in the way a circuit solves the "temporal credit assignment problem" inherent in most learning tasks, thereby providing a candidate synaptic mechanism for meta-learning.

neuroscience↗

bioRxiv2022

Interactions between circuit architecture and plasticity in a closed-loop system

Determining the sites of plasticity underlying changes in neural activity and behavior is critical for understanding mechanisms of learning. Identifying such sites from neural recording data can be challenging due to feedback pathways that impede reasoning about cause and effect. We studied interactions between feedback, neural activity, and plasticity in the context of a closed-loop motor learning task for which there is disagreement about the loci and directions of plasticity. We constructed a set of models that differed in the strength of their recurrent feedback. Despite these differences, each model successfully fit a large set of neural and behavioral data. However, the patterns of plasticity predicted by the models fundamentally differed, with the sign of plasticity at a key site changing from depression to potentiation as feedback strength increased. Guided by our analysis, we suggest how such models can be experimentally disambiguated. Our results address a long-standing debate regarding cerebellum-dependent motor learning and demonstrate how learning-related changes in neural activity can appear to contradict the sign of the underlying plasticity when feedback is present.

neuroscience↗

bioRxiv2021

Oculomotor plant and neural dynamics suggest gaze control requires integration on distributed timescales

A fundamental principle of biological motor control is that the neural commands driving movement must conform to the response properties of the motor plants they control. In the oculomotor system, characterizations of oculomotor plant dynamics traditionally supported models in which the plant responds to neural drive to extraocular muscles on exclusively short, subsecond timescales. These models predict that the stabilization of gaze during fixations between saccades requires neural drive that approximates eye position on longer timescales and is generated through the temporal integration of brief eye velocity-encoding signals that cause saccades. However, recent measurements of oculomotor plant behaviour have revealed responses on longer timescales. Furthermore, measurements of firing patterns in the oculomotor integrator have revealed a more complex encoding of eye movement dynamics. Yet, the link between these observations has remained unclear. Here we use measurements from the larval zebrafish to link dynamics in the oculomotor plant to dynamics in the neural integrator. The oculomotor plant in both anaesthetized and awake larval zebrafish was characterized by a broad distribution of response timescales, including those much longer than one second. Analysis of the firing patterns of oculomotor integrator neurons, which exhibited a broadly distributed range of decay time constants, demonstrates the sufficiency of this activity for stabilizing gaze given an oculomotor plant with distributed response timescales. This work suggests that leaky integration on multiple, distributed timescales by the oculomotor integrator reflects an inverse model for generating oculomotor commands, and that multi-timescale dynamics may be a general feature of motor circuitry. KEY POINTSO_LIRecent observations of oculomotor plant response properties and neural activity across the oculomotor system have called into question classical formulations of both the oculomotor plant and the oculomotor integrator. C_LIO_LIHere we use measurements from new and published experiments in the larval zebrafish together with modelling to reconcile recent oculomotor plant observations with oculomotor integrator function. C_LIO_LIWe developed computational techniques to characterize oculomotor plant responses over several seconds in awake animals, demonstrating that long timescale responses seen in anesthetized animals extend to the awake state. C_LIO_LIAnalysis of firing patterns of oculomotor integrator neurons demonstrates the sufficiency of this activity for stabilizing gaze given an oculomotor plant with multiple, distributed response timescales. C_LIO_LIOur results support a formulation of gaze stabilization by the oculomotor system in which commands for stabilizing gaze are generated through integration on multiple, distributed timescales. C_LI

neuroscience↗

bioRxiv2020

Modularity and neural coding from a brainstem synaptic wiring diagram

How much can connectomes with synaptic resolution help us understand brain function? An optimistic view is that a connectome is a major determinant of brain function and a key substrate for simulating a brain. Here we investigate the explanatory power of connectomics using a wiring diagram reconstructed from a larval zebrafish brainstem. We identify modules of strongly connected neurons that turn out to be specialized for different behavioral functions, the control of eye and body movements. We then build a neural network model using a synaptic weight matrix based on the reconstructed wiring diagram. This leads to predictions that statistically match the neural coding of eye position as observed by calcium imaging. Our work shows the promise of connectome-based brain modeling to yield experimentally testable predictions of neural activity and behavior, as well as mechanistic explanations of low-dimensional neural dynamics, a widely observed phenomenon in nervous systems.

neuroscience↗

bioRxiv2019

Choice-selective sequences in cortical inputs to the NAc provide a potential substrate for credit assignment

How are actions linked with subsequent outcomes to guide choices? The nucleus accumbens, which is implicated in this process, receives glutamatergic inputs from the prelimbic cortex and midline regions of the thalamus. However, little is known about whether and how representations differ across these input pathways. By comparing these inputs during a reinforcement learning task in mice, we discovered that prelimbic cortical inputs preferentially represent actions and choices, whereas midline thalamic inputs preferentially represent cues. Choice-selective activity in the prelimbic cortical inputs is organized in sequences that persist beyond the outcome. Through computational modeling, we demonstrate that these sequences can support the neural implementation of reinforcement learning algorithms, both in a circuit model based on synaptic plasticity, and one based on neural dynamics. Finally, we test and confirm predictions of our circuit models by direct manipulation of nucleus accumbens input neurons. Thus, we integrate experiment and modeling to suggest neural solutions for credit assignment.

neuroscience↗

bioRxiv2018

Unsupervised discovery of temporal sequences in high-dimensional datasets, with applications to neuroscience

Identifying low-dimensional features that describe large-scale neural recordings is a major challenge in neuroscience. Repeated temporal patterns (sequences) are thought to be a salient feature of neural dynamics, but are not succinctly captured by traditional dimensionality reduction techniques. Here we describe a software toolbox--called seqNMF--with new methods for extracting informative, non-redundant, sequences from high-dimensional neural data, testing the significance of these extracted patterns, and assessing the prevalence of sequential structure in data. We test these methods on simulated data under multiple noise conditions, and on several real neural and behavioral data sets. In hippocampal data, seqNMF identifies neural sequences that match those calculated manually by reference to behavioral events. In songbird data, seqNMF discovers neural sequences in untutored birds that lack stereotyped songs. Thus, by identifying temporal structure directly from neural data, seqNMF enables dissection of complex neural circuits without relying on temporal references from stimuli or behavioral outputs.

neuroscience↗

bioRxiv2017

Population-scale organization of cerebellar granule neuron signaling during a visuomotor behavior

Granule cells at the input layer of the cerebellum comprise over half the neurons in the human brain and are thought to be critical for learning. However, little is known about granule neuron signaling at the population scale during behavior. We used calcium imaging in awake zebrafish during optokinetic behavior to record transgenically identified granule neurons throughout a cerebellar population. A significant fraction of the population was responsive at any given time. In contrast to core precerebellar populations, granule neuron responses were relatively heterogeneous, with variation in the degree of rectification and the balance of excitation versus inhibition. Functional correlations were strongest for nearby cells, with weak spatial gradients in the degree of rectification and excitation. These data open a new window upon cerebellar function and suggest granule layer signals represent elementary building blocks underrepresented in core sensorimotor pathways, thereby enabling the construction of novel patterns of activity for learning.\n\nSIGNIFICANCE STATEMENTCerebellar processing is important for a variety of fine motor tasks and sensorimotor adaptations, and a growing body of evidence indicates a prominent role in cognitive control. However, it has been challenging to understand cerebellar function during behavior because of difficulties in recording from cerebellar granule neurons, the most populous neuron type in the brain. We use population-scale optical imaging in the larval zebrafish to compare precerebellar activity to granule cell signaling. Our results suggest a behaviorally relevant expansion of precerebellar signaling representations at the granule layer of the cerebellum.

neuroscience↗