Biology subjects

Pan-Vazquez, A.

Publications and source records attributed to Pan-Vazquez, A..

10 recordsLinked to original sources

bioRxiv2026

Evolution imposes an inductive bias that alters and accelerates learning dynamics

The learning dynamics of biological brains and artificial neural networks are of interest to both neuroscience and machine learning. A key difference between them is that neural networks are often trained from a randomly initialized state whereas each brain is the product of generations of evolutionary optimization, yielding innate structures that enable few-shot learning and inbuilt reflexes. Artificial neural networks, by contrast, require non-ethological quantities of training data to attain comparable performance. To investigate the effect of evolutionary optimization on the learning dynamics of neural networks, we combined algorithms simulating natural selection and online learning to produce a method for evolutionarily conditioning artificial neural networks, and applied it to both reinforcement and supervised learning contexts. We found the evolutionary conditioning algorithm, by itself, performs comparably to an unoptimized baseline. However, evolutionarily conditioned networks show signs of unique and latent learning dynamics, and can be rapidly fine-tuned to optimal performance. These results suggest evolution constitutes an inductive bias that tunes neural systems to enable rapid learning.

neuroscience↗

bioRxiv2025

VTA dopamine neuron activity produces spatially organized value representations

What is the neural architecture by which dopamine (DA) determines choice? Reinforcement learning (RL) has suggested an algorithmic chain: prediction errors based on reward train predicted values for stimuli and actions, and thereby determine choice. Elements of this chain have been identified, but have yet to be convincingly assembled. Here we used optogenetic stimulation of putative prediction error neurons at the top of this chain - i.e., DA neurons in the ventral tegmental area (VTA) - and recorded the activity of thousands of neurons in their target structure - the striatum - in mice performing a task that dissociates VTA DA signalling, stimulus value, action value, and choice. Conventional RL models captured some features of the data, including DA-driven choice reinforcement and value correlates, but failed on two key observations: i) action and state value correlates were relatively weak in the VTAs main target, the nucleus accumbens (NAc), and ii) VTA DA could not reinforce actions in the absence of a reward-predictive cue (CS+). Instead, we find that VTA DA produces stimulus value representations in the NAc, which in turn generates action value representations downstream (e.g. the CP). We formalize these observations in a Conditioned Reinforcer model, where VTA DA specifically acts on stimuli rather than actions, and these stimulus value representations are used to drive choice reinforcement.

neuroscience↗

bioRxiv2025

Striatal pathways oppositely shift cortical activity along the decision axis

The cortex and basal ganglia are organized into multiple parallel loops that serve motor, limbic, and cognitive functions. The classic model of cortico-basal ganglia interactions posits that within each loop, the direct pathway of the basal ganglia activates the cortex and the indirect pathway inhibits it1-3. While this model has found support in the motor domain4,5, whether opponent control by the two pathways extends to the cognitive domain remains unknown. Here, we record from anterior cingulate cortex (ACC) and dorsomedial striatum (DMS) while inhibiting direct or indirect pathway neurons in DMS, as mice perform an accumulation-of-evidence task6-10. Inconsistent with the classic model, the manipulations do not produce opponent changes in overall ACC activity. Instead, the pathways exert opponent influence over a subpopulation of ACC neurons that encode accumulated sensory evidence, the task-relevant decision variable. The direction of the modulation depends on a neurons tuning to ipsilateral versus contralateral evidence, such that the two pathways generate opponent shifts in coding specifically along the decision axis. Thus, our results uncover unexpected specificity in the effects of basal ganglia pathways on the cortex, with the two pathways of the DMS exerting opponent control not on overall activity but on coding of the relevant task variable. This functional specificity may extend to other basal ganglia loops to support different aspects of adaptive behavior, with the pathways serving a general role in selecting and shifting cortical representations to subserve circuit-specific functions.

neuroscience↗

bioRxiv2024

Interactive data exploration websites for large-scale electrophysiology

Methodological advances in neuroscience have enabled the collection of massive datasets which demand innovative approaches for scientific communication. Existing platforms for data storage lack intuitive tools for data exploration, limiting our ability to interact effectively with these brain-wide datasets. We introduce two public websites: (Data and Atlas) developed for the International Brain Laboratory which provide access to millions of behavioral trials and hundreds of thousands of individual neurons. These interfaces allow users to discover both the raw and processed brain-wide data released by the IBL at the scale of the whole brain, individual sessions, trials, and neurons. By hosting these data interfaces as websites they are available cross-platform with no installation. By releasing each sites code as a modular open-source framework, other researchers can easily develop their own web interfaces and explore their own data. As neuroscience datasets continue to expand, customizable web interfaces offer a glimpse into a future of streamlined data exploration and act as blueprints for future tools.

neuroscience↗

bioRxiv2023

Dissecting the Complexities of Learning With Infinite Hidden Markov Models

Learning to exploit the contingencies of a complex experiment is not an easy task for animals. Individuals learn in an idiosyncratic manner, revising their approaches multiple times as they are shaped, or shape themselves, and potentially end up with different strategies. Their long-run learning curves are therefore a tantalizing target for the sort of individualized quantitative characterizations that sophisticated modelling can provide. However, any such model requires a flexible and extensible structure which can capture radically new behaviours as well as slow changes in existing ones. To this end, we suggest a dynamic input-output infinite hidden semi-Markov model, whose latent states are associated with specific components of behaviour. This model includes an infinite number of potential states and so has the capacity to describe substantially new behaviours by unearthing extra states; while dynamics in the model allow it to capture more modest adaptations to existing behaviours. We individually fit the model to data collected from more than 100 mice as they learned a contrast detection task over tens of sessions and around fifteen thousand trials each. Despite large individual differences, we found that most animals progressed through three major stages of learning, the transitions between which were marked by distinct additions to task understanding. We furthermore showed that marked changes in behaviour are much more likely to occur at the very beginning of sessions, i.e. after a period of rest, and that response biases in earlier stages are not predictive of biases later on in this task.

animal behavior and cognition↗

bioRxiv2023

A neural mechanism for learning from delayed postingestive feedback

Animals learn the value of foods based on their postingestive effects and thereby develop aversions to foods that are toxic1-6 and preferences to those that are nutritious7-14. However, it remains unclear how the brain is able to assign credit to flavors experienced during a meal with postingestive feedback signals that can arise after a substantial delay. Here, we reveal an unexpected role for postingestive reactivation of neural flavor representations in this temporal credit assignment process. To begin, we leverage the fact that mice learn to associate novel15-18, but not familiar, flavors with delayed gastric malaise signals to investigate how the brain represents flavors that support aversive postingestive learning. Surveying cellular resolution brainwide activation patterns reveals that a network of amygdala regions is unique in being preferentially activated by novel flavors across every stage of the learning process: the initial meal, delayed malaise, and memory retrieval. By combining high-density recordings in the amygdala with optogenetic stimulation of genetically defined hindbrain malaise cells, we find that postingestive malaise signals potently and specifically reactivate amygdalar novel flavor representations from a recent meal. The degree of malaise-driven reactivation of individual neurons predicts strengthening of flavor responses upon memory retrieval, leading to stabilization of the population-level representation of the recently consumed flavor. In contrast, meals without postingestive consequences degrade neural flavor representations as flavors become familiar and safe. Thus, our findings demonstrate that interoceptive reactivation of amygdalar flavor representations provides a neural mechanism to resolve the temporal credit assignment problem inherent to postingestive learning.

neuroscience↗

bioRxiv2023

A Brain-Wide Map of Neural Activity during Complex Behaviour

A key challenge in neuroscience is understanding how neurons in hundreds of interconnected brain regions integrate sensory inputs with prior expectations to initiate movements and thereby make decisions. It is difficult to meet this challenge if different laboratories apply different analyses to different recordings in different regions during different behaviours. Here, we report a comprehensive set of recordings from 621733 neurons across 139 mice in 12 labs performing a decision-making task with sensory, motor, and cognitive components, obtained with 699 Neuropixels probe insertions covering 279 brain areas in the left forebrain and midbrain and the right hindbrain and cerebellum. We provide an initial appraisal of this brain-wide map, assessing how neural activity encoded key task variables. Representations of visual stimuli appeared transiently in classical visual areas after stimulus onset and then spread to ramp-like activity in a collection of mid- and hindbrain regions that also encoded choices. Neural responses correlated with motor action almost everywhere in the brain. Responses to reward delivery and consumption versus negative feedback were also widespread. This publicly available dataset represents an unprecedented resource for understanding how computations distributed across and within brain areas drive behaviour.

neuroscience↗

bioRxiv2023

Brain-wide representations of prior information in mouse decision-making

The neural representations of prior information about the state of the world are poorly understood. To investigate them, we examined brain-wide Neuropixels recordings and widefield calcium imaging collected by the International Brain Laboratory. Mice were trained to indicate the location of a visual grating stimulus, which appeared on the left or right with prior probability alternating between 0.2 and 0.8 in blocks of variable length. We found that mice estimate this prior probability and thereby improve their decision accuracy. Furthermore, we report that this subjective prior is encoded in at least 20% to 30% of brain regions which, remarkably, span all levels of processing, from early sensory areas (LGd, VISp) to motor regions (MOs, MOp, GRN) and high level cortical regions (ACAd, ORBvl). This widespread representation of the prior is consistent with a neural model of Bayesian inference involving loops between areas, as opposed to a model in which the prior is incorporated only in decision-making areas. This study offers the first brain-wide perspective on prior encoding at cellular resolution, underscoring the importance of using large scale recordings on a single standardized task.

neuroscience↗

bioRxiv2023

Lightning Pose: improved animal pose estimation via semi-supervised learning, Bayesian ensembling, and cloud-native open-source tools

Contemporary pose estimation methods enable precise measurements of behavior via supervised deep learning with hand-labeled video frames. Although effective in many cases, the supervised approach requires extensive labeling and often produces outputs that are unreliable for downstream analyses. Here, we introduce "Lightning Pose," an efficient pose estimation package with three algorithmic contributions. First, in addition to training on a few labeled video frames, we use many unlabeled videos and penalize the network whenever its predictions violate motion continuity, multiple-view geometry, and posture plausibility (semi-supervised learning). Second, we introduce a network architecture that resolves occlusions by predicting pose on any given frame using surrounding unlabeled frames. Third, we refine the pose predictions post-hoc by combining ensembling and Kalman smoothing. Together, these components render pose trajectories more accurate and scientifically usable. We release a cloud application that allows users to label data, train networks, and predict new videos directly from the browser.

neuroscience↗

bioRxiv2018

Homeostatic plasticity rules control the wiring of axo-axonic synapses at the axon initial segment

GABAergic interneurons are chiefly responsible for controlling the activity of local circuits in the cortex1,2. However, the rules that govern the wiring of interneurons are not well understood3. Chandelier cells (ChCs) are a type of GABAergic interneuron that control the output of hundreds of neighbouring pyramidal cells through axo-axonic synapses which target the axon initial segment (AIS)4. Despite their importance in modulating circuit activity, our knowledge of the development and function of axo-axonic synapses remains elusive. In this study, we investigated the role of activity in the formation and plasticity of ChC synapses. In vivo imaging of ChCs during development uncovered a narrow window (P12-P18) over which axons arborized and formed connections. We found that increases in the activity of either pyramidal cells or individual ChCs during this temporal window resulted in a reversible decrease in axo-axonic connections. Voltage imaging of GABAergic transmission at the AIS showed that axo-axonic synapses were depolarising during this period. Identical manipulations of network activity in older mice (P40-P46), when ChC synapses are inhibitory, resulted in an increase in axo-axonic synapses. We propose that the direction of ChC plasticity follows homeostatic rules that depend on the polarity of axo-axonic synapses.

neuroscience↗