Frontiers in Computational Neuroscience

Spike-timing-dependent plasticity (STDP) and dopamine (DA) are fundamental to reward-based learning and memory formation. A widely used DA-modulated STDP model explains how neural networks associate stimuli with delayed dopaminergic rewards through an eligibility trace. However, we show that this model supports learning even at unrealistically high DA concentrations because DA simply scales the magnitude of STDP without changing its temporal profile. In contrast, experiments demonstrate that DA nonlinearly reshapes the STDP window, converting long-term depression (LTD) into long-term potentiation (LTP) at high DA levels. We therefore propose a DA-modulated STDP rule in which increasing DA progressively biases plasticity toward potentiation while receptor saturation limits further DA effects beyond a critical concentration. Simulations of recurrent networks of Izhikevich neurons show that the proposed rule supports robust conditioning only within a biologically realistic DA range (0.04-0.70 {micro}M). Successful learning produces a hybrid network architecture consisting of a strong feedforward backbone embedded within recurrent circuitry and generates enhanced burst responses selectively to reward-associated stimuli. At the upper limit of the biologically plausible DA range, the network passes through a narrow bistable regime, converging to one of two distinct stable configurations. At higher DA concentrations, conditioning fails altogether. These results provide a biologically grounded model of DA-dependent plasticity and offer new insight into how abnormal dopamine signaling can impair learning in neurological disorders.

Equilibrium propagation (EP) is a biologically plausible alternative to backpropagation for training neural networks. EP typically relies on free and nudged dynamical phases: during the free phase, the network relaxes toward an equilibrium state, whereas during nudging, the output state is perturbed toward a target using a teaching signal. However, it remains unclear whether the brain has access to such explicit target signals. Inspired by Attention-Gated Brain Propagation (BrainProp), a reward-based learning framework proposed by Pozzi et al. (2020), we introduce a reward-based variant of EP that replaces full target-based nudging with a selected-output binary reward signal. The proposed method updates the network using only the chosen class and whether that choice is correct, without directly revealing the full target vector. We evaluate the method on MNIST, Fashion-MNIST, and CIFAR-10 using both multilayer perceptrons and convolutional neural networks. The proposed reward-based EP achieves performance close to that of conventional EP across all three datasets, although it generally converges more slowly during the early stages of training. Generalization-gap analyses show similar behavior for the two methods on MNIST and Fashion-MNIST, while reward-based EP exhibits a smaller training-test accuracy gap during later training on CIFAR-10. We further investigate the effect of the exploration probability used during stochastic class selection and find that moderate exploration can provide small performance improvements, although its effect is dataset-dependent. These results demonstrate that EP can learn effectively from sparse, action-specific reward feedback rather than a complete supervised target.

Animals must often navigate environments where feedback about progress toward a goal is sparse or delayed, requiring internal representations of space and memory of prior experience. The hippocampal-entorhinal system is believed to support this capability through distributed spatial representations that guide goal-directed behavior. However, many computational models of these circuits focus primarily on reproducing neural dynamics rather than demonstrating how such representations support learning on navigation tasks. We present a biologically inspired spiking neuronal network (SNN) model that combines grid-cell-derived spatial representations, {Delta}Q-modulated Hebbian plasticity, and context-dependent modulation to support navigation under sparse reward conditions. Grid Cell populations generate distributed spatial codes that are transformed by an Association Cell population into more spatially selective internal representations. Learning is driven by changes in Q-values ({Delta}Q) computed from a goal-conditioned Q-table, allowing local synaptic plasticity to incorporate information about long-term navigation outcomes. For environments containing multiple navigation objectives, a Context Cell population provides task-dependent modulation that enables a shared network architecture to support distinct navigation policies. Across two complementary maze environments, the model demonstrates three core capabilities: generation of distinct spatial representations, learning of efficient navigation policies under sparse and delayed reward, and support for multiple navigation objectives within a shared environment. The results further show that contextual modulation introduces subtle task-dependent variations into a largely shared population representation, allowing identical spatial locations to support different navigation behaviors. These findings demonstrate that biologically inspired spatial representations, value-guided plasticity, and contextual modulation can jointly support flexible navigation in spiking neuronal networks, providing a bridge between mechanistic neural circuit models and functional reinforcement learning.

Wave recurrent neural networks (wRNNs) are biologically inspired recurrent architectures that use traveling-wave dynamics to support sequence learning and memory. However, their input-to-hidden pathway remains relatively simple compared with biological neurons, where dendrites perform nonlinear input integration. In this study, we introduce the Dendritic Wave Recurrent Neural Network (DWRNN), which augments the input pathway of the wRNN with nonlinear basal dendritic branches while preserving the original recurrent wave dynamics. We evaluate DW-RNN on a simple copy task, sequential MNIST (sMNIST), permuted sequential MNIST (psMNIST), and noisy sequential CIFAR-10 (nsCIFAR-10). On the copy task, DW-RNN shows learning behavior comparable to the standard wRNN, suggesting that dendritic input integration does not disrupt the recurrent wave-based memory mechanism. On the three sequential image-classification benchmarks, DW-RNN outperforms the standard wRNN, improving accuracy from 97.27 {+/-} 0.15% to 97.82 {+/-} 0.12% on sMNIST, from 96.74 {+/-} 0.17% to 96.92 {+/-} 0.10% on psMNIST, and from 54.30 {+/-} 0.79% to 55.65 {+/-} 0.55% on nsCIFAR-10. In addition to improving mean accuracy, DW-RNN exhibits lower across-seed variability on all three classification benchmarks, suggesting that dendritic input integration may improve the stability of wRNN training. Hidden-activity visualizations further show that DW-RNN preserves the characteristic traveling-wave patterns of the original wRNN. These results suggest that dendritic computation and traveling-wave recurrent dynamics provide complementary mechanisms for biologically inspired sequence learning.

Neural synchronization is fundamental to brain function and, when it becomes excessive, underlies pathological conditions such as epilepsy. Among brain regions, the temporal lobes, and the hippocampus in particular, exhibit the highest epileptogenic potential, with mesial temporal lobe epilepsy representing the most prevalent form of the condition in humans. Within the hippocampus, extracellular potassium dynamics are central to non-synaptic epileptiform activity, and astrocytic potassium buffering mechanisms have emerged as key regulators of network excitability. Yet the specific contributions of astrocytic gap-junction coupling and potassium spatial buffering to neuronal synchronization across different spatial scales remain poorly understood. To address this gap, we developed a microcircuit biophysical model consisting of two astrocyte-neuron modules, each comprising one astrocyte coupled to five neurons. Astrocyte-neuron interactions are mediated exclusively through shared extracellular potassium dynamics. Using a reduced astrocyte model that captures both local membrane and syncytial potassium buffering, we systematically investigated how astrocytic potassium handling shapes neuronal activity patterns and inter-module synchronization. Our results demonstrate that astrocytes prevent the emergence of pathological states -- such as sustained ictal activity and depolarization block, by stabilizing extracellular potassium levels. Furthermore, we show that astrocytic gap-junction coupling strength critically regulates phase synchronization between neuronal modules: stronger coupling promotes inter-module synchrony under physiological conditions, whereas impaired astrocytic function drives networks toward pathological hypersynchronization when extracellular potassium is elevated. These findings support the hypothesis that astrocytic networks impose modularity on hippocampal neuronal assemblies, and suggest that astrocytic connexins may represent a relevant therapeutic target in epilepsy and other disorders characterized by aberrant neural synchronization. Author summary

This paper introduces a sensorimotor learning framework for a corticocerebellar network, grounded in the perspective of population dynamics. Using an optimal control theory approach, the cerebellum model enhances preparatory activity through premotor input, allowing the motor cortex to reach the desired initial conditions for movement more efficiently. Unlike traditional motor learning approaches that focus on acquiring new skills, this paradigm emphasizes automatization of already executable behaviors through repetition driven by intrinsic motivation. The proposed model is evaluated using a center-out reaching task, demonstrating that the role of the cerebellum is to shorten the preparatory period required for the successful execution of the movement. These findings suggest that corticocerebellar interactions play a crucial role in optimizing motor preparation, offering insight into the neural mechanisms underlying movement efficiency.

Cervical cancer represents a pressing global health challenge, emphasizing the critical need for accurate and timely diagnostic methods to facilitate effective treatment and improve survival rates. In response to this challenge, the study presents CerViX-Net, an innovative classification framework designed to advance cervical cancer detection through enhanced computational efficiency and diagnostic accuracy. The development of CerViX-Net is motivated by the limitations of traditional diagnostic models, particularly in handling the computational and memory demands of large-scale data, while ensuring precise feature extraction and classification. CerViX-Net employs a hybrid deep learning architecture that combines the capabilities of ResNet50, EfficientNet-B0, and a Modified Vision Transformer (ViT) module. The ResNet50 branch extracts hierarchical features through stacked convolutional and identity blocks. In another path, the modified ViT module transforms image patches via linear projection, augments them with positional and class embeddings, and processes them using Parallel Transformer Encoder layers to model contextual relationships. Concurrently, EfficientNet-B0 utilizes MBConv blocks to extract multi-scale representations. The feature outputs from all three branches are integrated and passed through a classification head consisting of dropout layers and dense layers to ensure robust and accurate predictions. The proposed framework is rigorously evaluated on the Mendeley LBC dataset, achieving exceptional performance metrics with an accuracy of 99.69%, precision of 99.28%, recall of 99.48%, and an F1-score of 99.52%. The robustness of CerViX-Net is further validated on the SIPaKMeD and Herlev Pap Smear datasets, where it demonstrates comparable excellence, underscoring its efficacy and adaptability across diverse cytology datasets. Statistical validation using Friedman's test further reinforces its superiority over competing methods.

Background / Introduction: Biologically inspired spiking neural networks can model adaptive behavior, but learning multiple goals is difficult because synaptic updates for different targets can interfere. We tested whether multi-timescale plasticity and context-specific credit assignment could improve continual multi-goal learning in a spiking navigation system inspired by entorhinal-hippocampal circuitry. Methods: We developed a closed-loop spiking model containing grid-like, place-like, target-related, association, and motor-output populations. An agent navigated in a two-dimensional environment with randomized starting locations and learned through reward-modulated spike-timing dependent plasticity (STDP/RL) and a novel evidence-gated plasticity (EGP) framework. EGP accumulates candidate synaptic modifications, evaluates them using reward evidence, and consolidates only changes that improve performance. A target-context variant maintained separate proposal stores and reward evaluation for each target. Results: STDP/RL learned and retained a single-target navigation policy, but multi-target training produced substantial interference, including attraction to incorrect targets after learning. Across 10 connectivity seeds, target-context EGP achieved higher late-stage reward than global EGP, improved weakest-target performance, and increased the fraction of targets achieving positive reward. In a longer continual-learning simulation, reward increased for all targets, TEST-phase performance increasingly exceeded TRAIN-phase performance, and proposal magnitudes grew over learning. Dwell-time confusion analyses showed that target-context EGP reduced wrong-target attraction and improved target selectivity relative to multi-target STDP/RL. Conclusions: These results demonstrate that spiking navigation circuits can learn goal-directed behavior using local plasticity, but robust multi-goal learning benefits from context-specific evidence-based consolidation. Target-context EGP provides a biologically motivated mechanism for reducing interference during continual reinforcement learning in spiking neural networks.

We have studied the emergence of slow relaxation oscillations in next generation neural mass models with spike frequency adaptation. Relaxation oscillations connect low firing state (Down state) to high firing state (Up state) via the slow adaptation. In the examined cases, the orbit relaxes towards the Up State via a sequence of collective damped oscillations (peaks of activity), thus revealing population bursting dynamics. The slower is the adaptation time scale the higher is the complexity (number of peaks) displayed by the relaxation oscillations. In particular, a chaos-induced spike-adding mechanism regulates the increase in the number of peaks. In analogy to what found in the Hidmarsh-Rose neuron model, two different types of chaotic behaviors have been identified: Population Spiking and Population Bursting Chaos. The increase of the adaptation strength leads to shorter (longer) Up (Down) state durations somehow mimicking the effect of charbachol in in vitro experiments, where spontaneous slow waves are observed. Indeed, the scenario depicted in [1], where an increase of the concentration of carbachol induces a transition from anesthesia-like to sleep-like dynamics is consistent with our results based on the variation of the adaptation strength. HighlightsO_LISpike Frequency Adaptation (SFA) promotes the emergence of Slow Relaxation Oscillations C_LIO_LISpike-adding mechanisms, controlled by SFA, lead to Relaxation Oscillations of increasing complexity C_LIO_LITwo types of chaotic behaviours: Population Spiking and Population Bursting Chaos C_LIO_LISFA regulates Up and Down States durations and their correlation C_LI

Serotonergic psychedelics, such as LSD, psilocybin, and DMT, have strong effects on human brain activity, yet their mechanisms of action are only partially understood. Here, we present a biophysically-based mean-field model that integrates cellular and network-level details to simulate the effects of these compounds at different spatial scales. By incorporating the brain-wide distribution of 5-HT2A receptors, our model mechanistically links receptor activation through reducing leak membrane potassium conductances, based on electrophysiological data. Our simulations reveal that this microscopic perturbation leads to the emergence of a brain state characterised by asynchronous irregular dynamics with increased firing rates and alterations in spectral power, in particular reduction of alpha-frequency bands, consistent with empirical findings. This change in dynamics is accompanied by an increase in spontaneous complexity, as quantified by the Lempel-Ziv complexity index, as observed experimentally. Furthermore, our model accurately replicates experimental findings regarding the Perturbational Complexity Index (PCI), demonstrating that PCI does not increase significantly by psychedelic drug administration. This crucial dissociation, where spontaneous complexity and spectral power are affected while perturbational complexity is preserved, highlights the distinct neurophysiological substrates underlying different metrics in psychedelic states. Our multiscale model provides a robust, mechanistic framework for understanding how psychedelics modulate global brain activity.

With 302 neurons and a rigorously characterized connectome, the nematode Caenorhabditis elegans represents a powerful model organism to study the fundamental roles of neuronal circuits in behavior. However, despite the breadth of research, many questions remain unanswered regarding how these organisms are able to successfully navigate their environment. Here, we present a biologically grounded dynamical circuit model for the investigation of sensory-guided behavior during C. elegans chemotaxis. Our mathematical model consists of the chemosensory neuron AWA, interneurons RIM and RIA, motor neurons, including SMDs and RMDs, and body wall muscles that provide proprioceptive feedback through stretch receptors. After optimization with an evolutionary algorithm, the model locomotes effectively toward a chemical attractant, realistically capturing nematode chemotactic behavior. Chemotaxis is ensured by sharp turns, which resemble the omega turns of living nematodes, as a key emergent property of the model. The sharp turning behavior is triggered by decreases in the concentration of the attractant. These result in reduced AWA activity, which in turn triggers disinhibition of RIM and subsequent changes in RIA oscillations. The ensuing coordinated changes in downstream motor neurons activity patterns produce sharp turns, which correct the nematodes path, so that the model worm heads toward the attractant, and remains at its proximity, after it reaches the gradient peak. The proposed framework, along with its emergent dynamics, provides new insights into the minimum requirements for C. elegans circuitry to display major features of its chemotactic behavior, including omega turns. In parallel, it generates experimentally testable hypotheses with respect to the participating neuronal elements.

Deep Brain Stimulation (DBS) is an established clinical treatment for a variety of neurological disorders, including Parkinsons Disease where it has been shown to reduce motor symptoms as well as disrupt pathological beta oscillations in the basal ganglia. The mechanisms of action of DBS on the collective activity of neuronal circuits is not fully understood. We use a recurrently-connected excitatory-inhbitory network based on the Brunel network architecture that can produce activity in a variety of states. Using a model of DBS that can reproduce observed effects such as antidromic activation, local somatic suppression, and axonal activation, we characterize the effect of stimulation across the entire parameter space of the network. We show that the effects of stimulation are dependent on the baseline state of the network, with the level of beta suppression dependent on the level of inhibition and the external drive. Specifically, networks with higher inhibition and lower drive show greater disruption of beta oscillations. We further show that networks in different states are preferentially sensitive to different frequencies of stimulation, suggesting that alternative protocols to the clinically standard high-frequency stimulation may have therapeutic efficacy.

Stabilizing synaptic plasticity together with the neurons activity has remained a central challenge in theoretical neuroscience since the introduction of Hebbian learning principles. Classical Hebbian learning rules typically lead to unbounded synaptic growth, motivating the development of stabilization mechanisms such as normalization methods, BCM-type learning, synaptic scaling and others. While these approaches can prevent divergence, they can also exhibit different limitations e.g. resulting in too-sparse synaptic configurations or leading to poor scalability with increasing network size. A recently introduced meta-plasticity mechanism, termed annealed linear learning (ALL), dynamically reduces the learning rate as neuronal output increases, thereby preserving stable and interpretable fixed-point behavior of the output. However, the original formulation leads to an irreversible decay of the learning rate, preventing adaptation to changing environmental conditions. To address this, in the present study, we balance learning rate reduction at large outputs with recovery at small outputs and in addition introduce forgetting that gradually reduces synaptic weights. These extensions allow the system to discard outdated representations and adapt to novel input conditions. Analytical investigations demonstrate that the favorable output fixed-point properties of the original ALL framework are preserved under the extended rule. Furthermore, simulations with an artificial agent show that the proposed mechanism enables robust and fast re-learning and adaptation in changing environments.

We present a novel astro-neural-field population model with application to migraine-related cortical spreading depolarization. The model is composed of four spatio-temporal state variables: excitatory and inhibitory membrane potentials, astrocytic potassium uptake recruitment, and extracellular potassium concentration. Extending a previous neural field model, we incorporate activity-dependent astrocytic potassium clearance via a nonlinear term coupled to astrocyte dynamics. The astrocyte transfer function, like its neural counterpart, exhibits three regimes governed by extracellular potassium, capturing its effect on clearance. This yields a more comprehensive framework, better fits experimental data, and provides new insights into the mechanisms of cortical spreading depolarization.

Recurrent neural networks (RNNs) trained with backpropagation through time (BPTT) use gradients and global error signals to solve tasks, while evolutionary algorithms (EAs) offer an alternative solution through gradient-free optimization. Both classes of methods can solve the same tasks by modifying network weights, but it remains unclear how the choice of training paradigm biases the final connectivity structure and resulting dynamics. When drawing conclusions from task-trained RNNs, especially as proxies for neurobiological computation, it is important to consider whether the resulting network structure is due to the training method itself. Here, we compare four training paradigms--BPTT, evolution strategies (ES), genetic algorithms (GA), and GA combined with Ojas Hebbian plasticity rule (GA+Oja) -- across RNNs of 32, 64, and 128 neurons. These training paradigms were applied to two contrasting tasks that are exemplary of tasks which animals perform: a discrete working memory task and a continuous sensorimotor integration task. For both tasks, we analyzed how each training algorithm modified the network through four measurements: weight change allocations across input (Win), recurrent (Wrec), and output (Wout) layers, the effective rank of the recurrent weight matrix, the dimensionality of hidden-state dynamics, and task accuracy. BPTT achieved near-perfect accuracy across all task conditions, progressively allocated more weight changes to Wout as the working memory tasks difficulty increased, and confined hidden-state activity to a lower-dimensional subspace than any evolutionary method. Evolutionary methods maintained higher recurrent effective rank, higher activity dimensionality, and no comparable difficulty-dependent reallocation toward the readout, all while maintaining comparable task accuracy as BPTT. These findings show how gradient-based and gradient-free algorithms discover distinct structural and dynamical solutions to the same computational problems, especially in tasks involving working memory, with important implications for analyzing task-trained RNNs as models of biological neural computation. Biological neural circuits, which are shaped by evolution and local plasticity rather than gradient descent, may operate in higher-dimensional regimes than gradient-trained RNN models predict.

The precise mechanism through which Deep Brain Stimulation (DBS) mitigates freezing episodes in Parkinsons disease remains unknown. We modelled Parkinsonian freezing using a model of the basal ganglia that adopts a Dynamic Neural Field network for simultaneous action selection (selecting which action is executed) and specification (resolving the continuous parameters governing how that action is performed). Action selection success was based on whether activation surpassed a threshold. The model robustly differentiated healthy and dopamine-depleted (Parkinsonian) conditions, producing stable action selection under healthy dopamine and impaired, freeze-prone dynamics following depletion. We then modelled DBS as a scaled reduction in afferent drive to the subthalamic nucleus and globus pallidus internus. While this formulation has succeeded in previous computational work, it did not yield consistent restoration of action selection within our framework, motivating future investigation of alternative DBS formulations.

Robotic motor control built on liquid neural networks and related continuous-time models, such as LTC and CfC, is typically trained offline via backpropagation through time and lacks an explicit mechanism for recalibrating online as plant dynamics change. We ask whether a frozen CfC core, whose liquid state spans a fixed continuous-time basis, can support cerebellar-style online adaptation by adapting only its linear readout with a climbing-fiber-like error signal. In a planar two-link reaching simulation with a velocity-dependent curl force field, we adapt the readout online with a feedback-error-learning (FEL) signal under a least-mean-squares (LMS) rule, leaving the core untouched. The frozen-core readout-only controller re-straightens curl-perturbed reaches and, upon field removal, produces a mirror-image after-effect, a behavioral signature consistent with internal-model learning, which a feedback-only controller does not produce. The result generalizes from a dense CfC to a sparse Neural-Circuit-Policy (NCP) wiring when the recurrent state, rather than the projected motor output, is used as the readout basis; it is robust to force-field strength and direction; and a recursive-least-squares variant adapts faster but de-adapts slowly because its covariance collapses, a rigidity that a covariance-reset safe-forgetting rule removes. Within the explored two-link planar simulation range, we did not find a readout-only failure case that required adapting the frozen core in the tested conditions. In this simulation study, adapting only the readout therefore provides a biologically inspired, low-cost online error-adaptation layer for offline-trained continuous-time controllers.

Mean-field models are extensively used in large-scale brain simulations because they provide a wieldy description of population dynamics while preserving key features of neural activity. Despite their widespread adoption, no common and reproducible methodology currently exists to systematically derive and validate mean-field models starting from biologically grounded single neuron dynamics. As a result, implementations are often ad hoc, difficult to reproduce and rarely reusable. Here we introduce BRIDGE, a modular, open-source Python pipeline that enables the bottom-up reconstruction, analysis, validation, and simulation of mean-field models from single neurons. The framework integrates single neurons modelling, network simulations, extraction of population statistics, parameters analysis, quantitative comparisons between spiking neural networks and corresponding mean-field representations, and simulation of network of mean-fields. Its flexible architecture allows users to incorporate different neuron models and to generate region-specific or state-dependent mean-field formulations. BRIDGE provides a reproducible foundation for developing biologically informed mean-field models suitable for large-scale and whole-brain simulations, supporting the transition from generic homogeneous population models toward region-specific ones. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=91 SRC="FIGDIR/small/742067v1_ufig1.gif" ALT="Figure 1"> View larger version (28K): org.highwire.dtl.DTLVardef@1124404org.highwire.dtl.DTLVardef@2f8b2aorg.highwire.dtl.DTLVardef@1598f37org.highwire.dtl.DTLVardef@c9814b_HPS_FORMAT_FIGEXP M_FIG C_FIG

Decision formation is commonly described as the accumulation of noisy evidence in a low-dimensional decision variable, but it remains unclear how this latent computation is implemented by heterogeneous neural responses. Here, we propose that ramping and sequentially firing neurons form complementary population codes for the same decision variable. Inspired by Laplace-domain neural representations of time, exponential receptive fields in a ramping population give rise to a translatable edge-like activity profile; localized receptive fields in a sequential population give rise to an aligned bump-like profile. We construct a continuous attractor neural network that dynamically maintains these complementary representations while implementing evidence accumulation along a shared latent manifold. At the behavioral level, simulations show that the circuit closely reproduces the single-trial trajectories, choice probabilities, and reaction-time statistics of a standard diffusion decision model while generating heterogeneous ramping and sequential neural responses. Our framework connects latent behavioral dynamics, population geometry, and recurrent circuit mechanisms. More broadly, it provides a circuit-level realization of computation in the Laplace domain that may support the representation and updating of continuous cognitive variables across decision making, timing, memory, and spatial cognition.

Recurrence is thought to enhance the robustness of biological vision, but how it achieves this feat is largely unknown. Perceptual robustness can be implemented through either lateral connections supporting local integration within a processing stage or feedback connections drawing on broader context from higher stages, and through either a discriminative objective optimising task-relevant classification or a generative objective learning to reconstruct the causes of visual input. But do these different types of recurrence engage distinct computational strategies? As this question is difficult to test in vivo, we endowed convolutional neural networks with varying recurrent architectures and training objectives, and evaluated the consequences for internal representations and behaviour across noise levels. Two distinct computational strategies emerged. Generative feedback followed a reductionist strategy, with representations becoming lower-dimensional through denoising, achieving robustness at moderate noise levels without noise training. Both discriminative lateral and feedback recurrence followed an expansionist strategy, increasing dimensionality to sharpen discriminability without denoising, but requiring noise training to achieve robustness. These dissociable signatures reflect fundamentally different computational mechanisms of robust vision and provide testable predictions for which form of recurrence the brain employs.