An official website of the United States government
In well-established first-order conditioning experiments, the concurrence of a sensory cue with reinforcement forms an association, allowing the cue to predict future reinforcement. In the insect mushroom body, a brain region central to learning and memory, such associations are encoded in the synapses between its intrinsic and output neurons. This process is mediated by the activity of dopaminergic neurons that encode reinforcement signals. In second-order conditioning, a new sensory cue is paired with an already established one that presumably activates dopaminergic neurons due to its predictive power of the reinforcement. We explored minimal circuit motifs in the mushroom body for their ability to support second-order conditioning using mechanistic models. We found that dopaminergic neurons can either be activated directly by the mushroom body’s intrinsic neurons or via feedback from the output neurons via several pathways. We demonstrated that the circuit motifs differ in their computational efficiency and robustness. Beyond previous research, we suggest an additional motif that relies on feedforward input of the mushroom body intrinsic neurons to dopaminergic neurons as a promising candidate for experimental evaluation. It differentiates well between trained and novel stimuli, demonstrating robust performance across a range of model parameters.
more »
« less
This content will become publicly available on December 1, 2026
Heterogeneous quantization regularizes spiking neural network activity
Abstract The learning and recognition of object features from unregulated input has been a longstanding challenge for artificial intelligence systems. Brains, on the other hand, are adept at learning stable sensory representations given noisy observations, a capacity mediated by a cascade of signal conditioning steps informed by domain knowledge. The olfactory system, in particular, solves a source separation and denoising problem compounded by concentration variability, environmental interference, and unpredictably correlated sensor affinities using a plastic network that requires statistically well-behaved input. We present a data-blind neuromorphic signal conditioning strategy, based on the biological system architecture, that normalizes and quantizes analog data into spike-phase representations, thereby transforming uncontrolled sensory input into a regular form with minimal information loss. Normalized input is delivered to a column of spiking principal neurons via heterogeneous synaptic weights; this gain diversification strategy regularizes neuronal utilization, yoking total activity to the network’s operating range and rendering internal representations robust to uncontrolled open-set stimulus variance. To dynamically optimize resource utilization while balancing activity regularization and resolution, we supplement this mechanism with a data-aware calibration strategy in which the range and density of the quantization weights adapt to accumulated input statistics.
more »
« less
- PAR ID:
- 10614701
- Publisher / Repository:
- Springer Nature
- Date Published:
- Journal Name:
- Scientific Reports
- Volume:
- 15
- Issue:
- 1
- ISSN:
- 2045-2322
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Intelligence involves processing sensory experiences into representations useful for prediction. Understanding sensory experiences and building these contextual representations without prior knowledge of sensor models and environment is a challenging unsupervised learning problem. Current machine learning methods process new sensory data using prior knowledge defined by either domain knowledge or datasets. When datasets are not available, data acquisition is needed, though automating exploration in support of learning is still an unsolved problem. Here we develop a method that enables agents to efficiently collect data for learning a predictive sensor model—without requiring domain knowledge, human input, or previously existing data—using ergodicity to specify the data acquisition process. This approach is based entirely on data-driven sensor characteristics rather than predefined knowledge of the sensor model and its physical characteristics. We learn higher quality models with lower energy expenditure during exploration for data acquisition compared to competing approaches, including both random sampling and information maximization. In addition to applications in autonomy, our approach provides a potential model of how animals use their motor control to develop high quality models of their sensors (sight, sound, touch) before having knowledge of their sensor capabilities or their surrounding environment.more » « less
-
Abstract During development, cortical activity is organized into distributed modular patterns that are a precursor of the mature columnar functional architecture. Theoretically, such structured neural activity can emerge dynamically from local synaptic interactions through a recurrent network with effective local excitation with lateral inhibition (LE/LI) connectivity. Utilizing simultaneous widefield calcium imaging and optogenetics in juvenile ferret cortex prior to eye opening, we directly test several critical predictions of an LE/LI mechanism. We show that cortical networks transform uniform stimulations into diverse modular patterns exhibiting a characteristic spatial wavelength. Moreover, patterned optogenetic stimulation matching this wavelength selectively biases evoked activity patterns, while stimulation with varying wavelengths transforms activity towards this characteristic wavelength, revealing a dynamic compromise between input drive and the network’s intrinsic tendency to organize activity. Furthermore, the structure of early spontaneous cortical activity – which is reflected in the developing representations of visual orientation – strongly overlaps that of uniform opto-evoked activity, suggesting a common underlying mechanism as a basis for the formation of orderly columnar maps underlying sensory representations in the brain.more » « less
-
Abstract Bio-inspired robot controllers are becoming more complex as we strive to make them more robust to, and flexible in, noisy, real-world environments. A stable heteroclinic network (SHN) is a dynamical system that produces cyclical state transitions using noisy input. SHN-based robot controllers enable sensory input to be integrated at the phase-space level of the controller, thus simplifying sensor-integrated, robot control methods. In this work, we investigate the mechanism that drives branching state trajectories in SHNs. We liken the branching state trajectories to decision-splits imposed into the system, which opens the door for more sophisticated controls -- all driven by sensory input. This work provides guidelines to systematically define an SHN topology, and increase the rate at which desired decision states in the topology are chosen. Ultimately, we are able to control the rate at which desired decision states activate for input signal-to-noise ratios across six orders of magnitude.more » « less
-
Abstract The striatum plays an important role in learning, selecting, and executing actions. As a major input hub of the basal ganglia, it receives and processes a diverse array of signals related to sensory, motor, and cognitive information. Aberrant neural activity in this area is implicated in a wide variety of neurological and psychiatric disorders. It is therefore important to understand the hallmarks of disrupted striatal signal processing. This review surveys literature examining howin vivostriatal microcircuit dynamics are impacted in animal models of one of the most widely studied movement disorders, Parkinson's disease. The review identifies four major features of aberrant striatal dynamics: altered relative levels of direct and indirect pathway activity, impaired information processing by projection neurons, altered information processing by interneurons, and increased synchrony.more » « less
