Biology subjects

Chalk, M.

Publications and source records attributed to Chalk, M..

11 recordsLinked to original sources

bioRxiv2026

The multi-scale geometry of neural encoding

As recorded neural populations become larger and artificial networks more powerful, there is growing interest in understanding and comparing what different populations encode. Representational geometry offers a powerful way to characterise this: it remaps stimulus space to stretch well-encoded stimulus differences while shrinking poorly encoded ones. However, this remapping can be constructed in many ways from neural responses. Here, we introduce a multi-scale representational geometry, derived from first principles, that captures how neurons encode stimulus differences at every scale. It is exactly related to mutual information, so independent of decoding assumptions. The more a stimulus difference contributes to information, the more the geometry stretches it. Applied to spatial, motor and visual codes, it shows how neurons combine to encode information, how areas differ, and which features are best encoded. A single framework thus maps the structure of encoded information, uniting geometric and information-theoretic views of neural coding.

neuroscience↗

bioRxiv2026

Primitives for motion segmentation in the retina

The center surround structure of ganglion cells receptive fields is suited for edge detection, a first step towards image segmentation. However, in a dynamical visual scene with moving, textured objects, it is less clear how the retina represents the local boundaries of these objects. Here, we show that the spatial selectivity of ganglion cells changes during their responses to a moving object. Specific cell types only respond to contrast changes near the edges of the moving object, while being insensitive to changes in other parts of their receptive field. Using a non-linear model to reproduce this result, we could isolate the mechanism responsible for this selective representation. These types of ganglion cells represent selectively textures only when they are near moving edges, and may thus provide useful primitives for motion segmentation.

neuroscience↗

bioRxiv2025

Stimulus sensitivity in noisy neural systems

Understanding how neural populations encode sensory information requires a precise definition of neuronal sensitivity to stimuli. While tuning curves and firing rates offer intuitive insights, theoretical frameworks based on signal-to-noise ratio and decoding efficiency identify Fisher information as the canonical measure of sensitivity, due to its relation with decoding performance. However, this relation holds only under restrictive conditions--when many neurons encode the same stimulus feature or when neural noise is weak. In realistic settings, such as when complex or high-dimensional stimuli are represented by a small ensemble of neurons, Fisher information becomes ill-defined or misleading. To overcome these limitations, we investigate two complementary information-theoretic quantities--the stimulus-specific information (ISSI) and the local information (Iloc)--and propose them as robust alternatives for quantifying sensitivity. We show that ISSI and Iloc converge with Fisher information when signal to noise is large, yet remain meaningful and interpretable beyond that regime. Importantly, these measures capture distinct aspects of sensitivity: ISSI quantifies how observing a response reduces stimulus uncertainty, whereas Iloc reflects how small stimulus perturbations reshape the systems posterior beliefs. Together, they offer a unifying perspective linking information-theoretic and statistical notions of sensitivity, bridging theoretical analysis and experimental investigation of neural coding.

neuroscience↗

bioRxiv2025

Direction-selective retinal ganglion cells encode motion direction uniformly, despite having discretely distributed cardinal preferences

On-Off direction-selective retinal ganglion cells (DS RGCs) exhibit broad tuning curves, responding robustly to motion aligned with or near their preferred direction. These cells comprise four major subtypes, each tuned to motion along one of the four cardinal axes: nasal, superior, temporal, or inferior. However, natural stimuli can move in any direction, and it remains unclear whether intermediate directions are encoded less effectively, or whether this cardinal organization nevertheless supports uniform direction encoding. Here, we combined previous electrophysiological recordings with an information-theoretic measure, the Stimulus Specific Information, to estimate the directional sensitivity of small populations of recorded neurons. This analysis revealed that DS RGC populations are uniformly sensitive across all directions of motion. We then asked whether the observed homogeneous sensitivity was a consequence of DS RGCs maximizing the average stimulus information. Simulations with artificially modified tuning curve widths revealed that DS cells prioritize avoiding pronounced drops in sensitivity over maximizing the average transmitted Information. Maximizing this minimal sensitivity may therefore be a principle to understand the organization of sensory systems.

neuroscience↗

bioRxiv2023

Signatures of criticality in efficient coding networks

The critical brain hypothesis states that the brain can benefit from operating close to a second-order phase transition. While it has been shown that several computational aspects of sensory information processing (e. g., sensitivity to input) are optimal in this regime, it is still unclear whether these computational benefits of criticality can be leveraged by neural systems performing behaviorally relevant computations. To address this question, we investigate signatures of criticality in networks optimized to perform efficient encoding of stimuli. We consider a spike-coding network of leaky integrate-and-fire neurons with synaptic transmission delays and input noise. Previously, it was shown that the performance of such networks varies non-monotonically with the noise amplitude. Interestingly, we find that in the vicinity of the optimal noise level for efficient coding, the network dynamics exhibits signatures of criticality, namely, the distribution of avalanche sizes follows a power law. When the noise amplitude is too low or too high for efficient coding, the network appears either super-critical or sub-critical, respectively. Our work suggests that two influential, and previously disparate theories of neural processing optimization - efficient coding, and criticality - may be intimately related.

neuroscience↗

bioRxiv2023

Scalable gaussian process inference of neural responses to natural images

1Predicting the responses of sensory neurons is a long-standing neuroscience goal. However, while there has been much progress in modeling neural responses to simple and/or artificial stimuli, predicting responses to natural stimuli remains an ongoing challenge. One the one hand, deep neural networks perform very well on certain data-sets, but can fail when data is limited. On the other hand, gaussian processes (GPs) perform well on limited data, but are generally poor at predicting responses to high-dimensional stimuli, such as natural images. Here we show how structured priors, e.g. for local and smooth receptive fields, can be used to scale up GPs to high-dimensional stimuli. We show that when we do this, a GP model largely outperforms a deep neural network trained to predict retinal responses to natural images, with largest differences observed when both models are trained on a very small data-set. Further, since GPs compute the uncertainty in their predictions, they are well-suited to closed-loop experiments, where stimuli are chosen actively so as to collect informative neural data. We show how this can be done in practice on our retinal data-set, so as to: (i) efficiently learn a model of retinal responses to natural images, using little data, and (ii) rapidly distinguish between competing models (e.g. a linear vs a non-linear model). In the future, our approach could be applied to other low-level sensory areas, beyond the retina.

neuroscience↗

bioRxiv2022

Encoding surprise by retinal ganglion cells

The efficient coding hypothesis posits that early sensory neurons transmit maximal information about sensory stimuli, given internal constraints. A central prediction of this theory is that neurons should preferentially encode stimuli that are most surprising. Previous studies suggest this may be the case in early visual areas, where many neurons respond strongly to rare or surprising stimuli. For example, previous research showed that when presented with a rhythmic sequence of full-field flashes, many retinal ganglion cells (RGCs) respond strongly at the instance the flash sequence stops, and when another flash would be expected. This phenomenon is called the omitted stimulus response. However, it is not known whether the responses of these cells varies in a graded way depending on the level of stimulus surprise. To investigate this, we presented retinal neurons with extended sequences of stochastic flashes. With this stimulus, the surprise associated with a particular flash/silence, could be quantified analytically, and varied in a graded manner depending on the previous sequences of flashes and silences. Interestingly, we found that RGC responses could be well explained by a simple normative model, which described how they optimally combined their prior expectations and recent stimulus history, so as to encode surprise. Further, much of the diversity in RGC responses could be explained by the model, due to the different prior expectations that different neurons had about the stimulus statistics. These results suggest that even as early as the retina many cells encode surprise, relative to their own, internally generated expectations.

neuroscience↗

bioRxiv2021

Human-in-the-loop optimization of visual prosthetic stimulation

AO_SCPLOWBSTRACTC_SCPLOWRetinal prostheses are a promising strategy to restore sight to patients with retinal degenerative diseases. These devices compensate for the loss of photoreceptors by electrically stimulating neurons in the retina. Currently, the visual function that can be recovered with such devices is very limited. This is due, in part, to current spread, unintended axonal activation, and the limited resolution of existing devices. Here we show, using a recent model of prosthetic vision, that optimizing how visual stimuli are encoded by the device can help overcome some of these limitations, leading to dramatic improvements in visual perception. We propose a strategy to do this in practice, using patients feedback in a visual task. The main challenge of our approach comes from the fact that, typically, one only has access to a limited number of noisy responses from patients. We propose two ways to deal with this: first, we use a model of prosthetic vision to constrain and simplify the optimization; second, we use preferential Bayesian optimization to efficiently learn the encoder using minimal trials. To test our approach, we presented healthy subjects with visual stimuli generated by a recent model of prosthetic vision, to replicate the perceptual experience of patients fitted with an implant. Our optimization procedure led to significant and robust improvements in perceived image quality, that transferred to increased performance in other tasks. Importantly, our strategy is agnostic to the type of prosthesis and thus could readily be implemented in existing implants.

neuroscience↗

bioRxiv2019

Inferring the function performed by a recurrent neural network

A central goal in systems neuroscience is to understand the functions performed by neural circuits. Previous top-down models addressed this question by comparing the behaviour of an ideal model circuit, optimised to perform a given function, with neural recordings. However, this requires guessing in advance what function is being performed, which may not be possible for many neural systems. To address this, we propose a new framework for optimising a recurrent network using multi-agent reinforcement learning (RL). In this framework, a reward function quantifies how desirable each state of the network is for performing a given function. Each neuron is treated as an agent, which optimises its responses so as to drive the network towards rewarded states. Three applications follow from this. First, one can use multi-agent RL algorithms to optimise a recurrent neural network to perform diverse functions (e.g. efficient sensory coding or motor control). Second, one could use inverse RL to infer the function of a recorded neural network from data. Third, the theory predicts how neural networks should adapt their dynamics to maintain the same function when the external environment or network structure changes. This could lead to theoretical predictions about how neural network dynamics adapt to deal with cell death and/or varying sensory stimulus statistics.

neuroscience↗

bioRxiv2018

Population model learned on different stimulus ensembles predicts network responses in the retina

Correlations in sensory neural networks have both extrinsic and intrinsic origins. Extrinsic or stimulus correlations arise from shared inputs to the network, and thus depend strongly on the stimulus ensemble. Intrinsic or noise correlations reflect biophysical mechanisms of interactions between neurons, which are expected to be robust to changes of the stimulus ensemble. Despite the importance of this distinction for understanding how sensory networks encode information collectively, no method exists to reliably separate intrinsic interactions from extrinsic correlations in neural activity data, limiting our ability to build predictive models of the network response. In this paper we introduce a general strategy to infer population models of interacting neurons that collectively encode stimulus information. The key to disentangling intrinsic from extrinsic correlations is to infer the couplings between neurons separately from the encoding model, and to combine the two using corrections calculated in a mean-field approximation. We demonstrate the effectiveness of this approach on retinal recordings. The same coupling network is inferred from responses to radically different stimulus ensembles, showing that these couplings indeed reflect stimulus-independent interactions between neurons. The inferred model predicts accurately the collective response of retinal ganglion cell populations as a function of the stimulus.

neuroscience↗

bioRxiv2017

Towards a unified theory of efficient, predictive and sparse coding

A central goal in theoretical neuroscience is to predict the response properties of sensory neurons from first principles. Several theories have been proposed to this end. \"Efficient coding\" posits that neural circuits maximise information encoded about their inputs. \"Sparse coding\" posits that individual neurons respond selectively to specific, rarely occurring, features. Finally, \"predictive coding\" posits that neurons preferentially encode stimuli that are useful for making predictions. Except in special cases, it is unclear how these theories relate to each other, or what is expected if different coding objectives are combined. To address this question, we developed a unified framework that encompasses these previous theories and extends to new regimes, such as sparse predictive coding. We explore cases when different coding objectives exert conflicting or synergistic effects on neural response properties. We show that predictive coding can lead neurons to either correlate or decorrelate their inputs, depending on presented stimuli, while (at low-noise) efficient coding always predicts decorrelation. We compare predictive versus sparse coding of natural movies, showing that the two theories predict qualitatively different neural responses to visual motion. Our approach promises a way to explain the observed diversity of sensory neural responses, as due to a multiplicity of functional goals performed by different cell types and/or circuits.

neuroscience↗