Biology subjects

Kar, K.

Publications and source records attributed to Kar, K..

At least 19 recordsLinked to original sources

bioRxiv2026

Motion-Dependent Object Perception Reveals Limits of Current Video Neural Networks

How does motion contribute to robust object perception when appearance cues are unreliable? In natural scenes, camouflage, clutter, and occlusion can obscure object boundaries in static images, yet humans often resolve these ambiguities once objects move. Here we ask whether modern artificial vision systems capture this motion-dependent computation and whether their internal representations align with those used by biological vision. Using videos from the MOCA (Moving Camouflaged Animals) dataset, we introduce behavioral benchmarks that quantify the accuracy of object position and size estimation in scenes containing either static or moving objects. We first compare human observers with a diverse set of artificial neural networks. For static stimuli, image-based and video-based models achieve similar accuracy in predicting object position and size. However, humans show systematic improvements when objects are presented in motion. Image-based models do not exhibit this motion-dependent improvement, whereas several video-based architectures reproduce this behavioral pattern by integrating information across time. To examine the representational basis of these differences, we record neural population responses from the macaque inferior temporal (IT) cortex during presentation of the same stimuli. Models that more closely match IT representations also better reproduce human motion-dependent behavior. These results show that static accuracy alone is insufficient to evaluate models of visual perception and that alignment with primate visual representations provides a useful guide for developing models that capture dynamic computations in vision.

neuroscience↗

bioRxiv2025

Adaptive recruitment of cortex-wide recurrence for visual object recognition

Theories of the neural mechanism underpinning rapid recognition debate whether it relies solely on a feedforward sweep through the ventral stream or instead requires recurrent processing, possibly engaging additional brain regions. Here we directly tested the "adaptive recurrence hypothesis", that attempts to unify these disparate views by proposing that additional recurrent cortical resources beyond the visual stream are recruited when feedforward processing alone is insufficient to solve object recognition. To investigate this hypothesis, we contrasted functional MRI (fMRI) and electroencephalography (EEG) responses to compare neural responses to images that are equally well recognized by humans, but that differ in whether they could be solved by a feedforward deep neural network; a computational proxy for ventral stream feedforward processing. We found that when feedforward processing in the ventral visual stream is insufficient, additional parieto-frontal networks are rapidly and transiently recruited, representationally reconfiguring the ventral visual stream. Our results reveal that object recognition flexibly adapts through fast, cortex-wide recurrence, providing a unifying framework for competing theories of visual recognition.

neuroscience↗

bioRxiv2025

Distinct contributions of memorability and object recognition to the representational goals of the macaque inferior temporal cortex

The primate inferior temporal (IT) cortex, at the apex of the ventral visual stream, encodes information that supports diverse representational goals--from recognizing objects to determining which images are likely to be remembered. Specific artificial neural networks (ANNs), that currently serve as the leading computational hypotheses of ventral stream processing, are typically trained exclusively for object recognition. We asked whether incorporating image memorability as an additional optimization objective could improve ANN-brain alignment. Models optimized for memorability explained additional, non-overlapping variance in IT responses beyond that captured by recognition-optimized networks, indicating that memorability and recognition rely on partly independent dimensions of IT representation. Notably, these models also exhibited fewer non-brain-like units, bringing their representational geometry closer to that of IT. Furthermore, networks jointly optimized for both objectives were more predictive of human memorability than memorability-only models, while maintaining their alignment with human object recognition performance patterns. Together, these findings suggest that IT encodes multiple representational goals and that models trained solely for recognition provide an incomplete account of ventral stream computation. SignificanceBrain regions often serve multiple representational goals, and identifying those goals is critical because they provide the key to building better encoding models of the system. The primate ventral visual stream has traditionally been understood as a pathway for object recognition, with the inferior temporal (IT) cortex regarded as its core substrate. However, IT responses also predict image memorability--a robust phenomenon whereby some images are consistently remembered better than others. Here we show that memorability constitutes a separable representational goal of IT. ANNs optimized for memorability explained neural variance not captured by recognition models, and the two objectives produced distinct representational geometries. Critically, models jointly optimized for both recognition and memorability provided the best match to IT responses, improved prediction of human memorability, and preserved recognition performance. These findings highlight memorability as an organizing principle of IT and demonstrate that multi-goal optimization yields more brain-like computational models of vision.

neuroscience↗

bioRxiv2025

Facial expression discrimination emerges from neural subspaces shared with detection and identity

Understanding how the human brain decodes facial expressions remains a fundamental challenge, requiring computational models that tightly connect neural responses to behavior. Here, we demonstrate that rhesus macaques provide a unique and powerful animal model to uncover the neural computations behind human facial expression discrimination, bridging critical gaps between behavior, neural activity, and computational theory. Despite the challenges of establishing reliable behavioral paradigms in macaques, we developed a robust discrimination task spanning six emotional categories, yielding strong, image-by-image behavioral correspondence between macaques and humans. By systematically comparing artificial neural networks (ANNs) to macaque behavior and IT neural data, we found that traditional action unit-based models fail to capture image-level behavioral structure, while ANNs with IT-like internal representations outperform all others. Neural recordings showed that the specific IT population responses (70-100 ms) carried the strongest predictive power for facial expression discrimination, underscoring the primacy of feedforward codes in guiding behavior. Expression coding in IT was significantly shaped by face-selective neurons that also encoded identity. This convergence points to a shared functional subspace in IT, where stable (identity) and dynamic (expression) information coexist along overlapping dimensions. Such an architecture moves beyond the classical view of segregated pathways, revealing a general coding principle by which IT flexibly supports multiple socially relevant functions within a common representational geometry.

neuroscience↗

bioRxiv2025

Reverse Predictivity: Going Beyond One-Way Mapping to Compare Artificial Neural Network Models and Brains

A major goal in systems neuroscience is to build computational models that capture the primate brains internal representations. Standard evaluations of artificial neural networks (ANNs) emphasize forward predictivity--how well model features predict neural responses--without testing whether model representations are themselves recoverable from neural activity. Here we develop the reverse predictivity metric, which quantifies how well macaque inferior temporal (IT) cortex responses predict ANN unit activations. This two-way framework reveals a striking asymmetry: models with high forward predictivity ([~]50% variance explained) often contain units unpredictable from neural activity, reflecting biologically inaccessible dimensions. In contrast, monkey-to-monkey mappings are symmetric, confirming that the asymmetry reflects genuine representational mismatch. Reverse predictivity isolates "common" ANN units--shared with IT, behaviorally relevant, and generalizing across species--and "unique" units lacking such alignment. Influenced by feature dimensionality, training objectives, and adversarial robustness, reverse predictivity offers a principled benchmark for guiding next-generation ANNs toward both high task performance and genuine biological plausibility.

neuroscience↗

bioRxiv2024

The effects of object category training on the responses of macaque inferior temporal cortex are consistent with performance-optimizing updates within a visual hierarchy

How does the primate brain coordinate plasticity when learning to discriminate new objects? We measured consequences of object learning on macaque inferior temporal (IT) cortex, a key waypoint supporting object recognition in the ventral visual stream. Neural activity in task-trained monkeys IT showed increased object selectivity, enhanced linear separability across objects, and more object-invariant representations compared to task-naive monkeys. To model these differences, we developed a computational framework using anatomically-mapped artificial neural network (ANN) models of the ventral stream with various learning algorithms. Simulations revealed that gradient-based, performance-optimizing updates of ANN internal representations accurately approximated observed IT cortex changes. These models predict novel training-induced phenomena in IT cortex, including changes independent of object identity and ITs alignment with behavior. This convergence between empirical measurements and model predictions suggests ventral stream plasticity follows task optimization principles well-approximated by gradient descent, enabling accurate predictions about visual plasticity and generalization to test images.

neuroscience↗

bioRxiv2024

The Impact of Scene Context on Visual Object Recognition: Comparing Humans, Monkeys, and Computational Models

During natural vision, we rarely see objects in isolation but rather embedded in rich and complex contexts. Understanding how the brain recognizes objects in natural scenes by integrating contextual information remains a key challenge. To elucidate neural mechanisms compatible with human visual processing, we need an animal model that behaves similarly to humans, so that inferred neural mechanisms can provide hypotheses relevant to the human brain. Here we assessed whether rhesus macaques could model human context-driven object recognition by quantifying visual object identification abilities across variations in the amount, quality, and congruency of contextual cues. Behavioral metrics revealed strikingly similar context-dependent patterns between humans and monkeys. However, neural responses in the inferior temporal (IT) cortex of monkeys that were never explicitly trained to discriminate objects in context, as well as current artificial neural network models, could only partially explain this cross-species correspondence. The shared behavioral variance unexplained by context-naive neural data or computational models highlights fundamental knowledge gaps. Our findings demonstrate an intriguing alignment of human and monkey visual object processing that defies full explanation by either brain activity in a key visual region or state-of-the-art models.

neuroscience↗

bioRxiv2024

Object motion representation in the macaque ventral stream -- a gateway to understanding the brain's intuitive physics engine

Effective interaction with moving objects and the ability to infer and predict their motion (a core component of "intuitive physics") is essential for survival in the dynamic world. How does the primate visual system process such stimuli, enabling predictive capabilities for dynamic stimuli statistics like motion velocity and expected trajectories? In this study, we probed brain areas in the ventral visual pathway of rhesus macaques implicated in object recognition (areas V4 and inferior temporal, IT, cortex) to evaluate how they represent object motion speed and direction. We assessed the relationship between the distributed population activity in the ventral stream and two distinct object motion-based behaviors--one reliant on information directly available in videos (speed discrimination) and the other predicated on predictive motion estimates from videos (future event predictions). Further, employing microstimulation strategies, we confirm the causal, functional role of the IT cortex in these behaviors. Our results underscore the need to re-examine the traditional functional segregation of the primate visual cortices into "what" and "where" pathways and provide empirical constraints to model their interaction for a better circuit-level understanding of visual motion and intuitive physics.

neuroscience↗

bioRxiv2023

Distinct roles of putative excitatory and inhibitory neurons in the macaque inferior temporal cortex in core object recognition behavior

A spatially distributed population of neurons in the macaque inferior temporal (IT) cortex supports object recognition behavior, but the cell-type specificity of the population in forming "behaviorally sufficient" object decodes remains poorly understood. To address this, we recorded neural signals from the macaque IT cortex and compared the object identity information and the alignment of decoding strategies derived from putative inhibitory (Inh) and excitatory (Exc) neurons to the monkeys behavior. We observed that while Inh neurons represented significant category information, decoding strategies based on Exc neural population activity outperformed those from Inh neurons in overall accuracy and their image-level match to the monkeys behavioral reports. Interestingly, both Exc and Inh responses explained a fraction of unique variance of the monkeys behavior, demonstrating a distinct role of the two cell types in generating object identity solutions for a downstream readout. We observed that current artificial neural network (ANN) models of primate ventral stream, designed with AI goals of performance optimization on image categorization, better predict Exc neurons (and their contribution to object recognition behavior) than Inh neurons. Beyond, refining the linking propositions between IT neurons and object recognition behavior, our results guide the development of next-generation biologically constrained brain models by offering novel cell-type specific neural benchmarks.

neuroscience↗

bioRxiv2022

An efficient method to generate fluorescent amyloid fibrils

Studies on fluorophore-tagged peptides help in elucidating the molecular mechanism of amyloidogenesis including their cellular internalization and crosstalk potential. Despite several advantages, unavoidable difficulties including expensive and tedious synthesis-protocols exist in fluorophore-based tools. Importantly, covalently-tagged fluorophores could introduce structural constraints which may influence the conformation of the monomeric and aggregated forms of protein. To resolve this problem, we describe a robust yet simple method to make fluorescent amyloid fibrils through non-covalent incorporation of fluorophores into amyloid fibrils. We used aggregation protocol in which a small amount of fluorophore is incorporated into the amyloids, and this protocol does not alter the aggregation kinetics and the characteristic {beta}-sheet-conformers of the generated amyloid fibrils. We have successfully prepared fluorescent amyloid fibrils of Insulin, Lysozyme and A{beta}1-42, and the noncovalently incorporated fluorophores remained intact in the amyloid fibrils without leaching, even after serial-dilutions and prolonged-storage. Further, this method enables successful monitoring of cellular-internalization of the fluorescent amyloids into SH-SY5Y and A549 cells, and it also detects FRET-signals during interfibrillar interactions. The findings establish a simple and affordable protocol to prepare fluorescent amyloid structures, which may significantly help amyloid researchers working on both in vitro and animal model systems.

biophysics↗

bioRxiv2022

Aligning Model and Macaque Inferior Temporal Cortex Representations Improves Model-to-Human Behavioral Alignment and Adversarial Robustness

While some state-of-the-art artificial neural network systems in computer vision are strikingly accurate models of the corresponding primate visual processing, there are still many discrepancies between these models and the behavior of primates on object recognition tasks. Many current models suffer from extreme sensitivity to adversarial attacks and often do not align well with the image-by-image behavioral error patterns observed in humans. Previous research has provided strong evidence that primate object recognition behavior can be very accurately predicted by neural population activity in the inferior temporal (IT) cortex, a brain area in the late stages of the visual processing hierarchy. Therefore, here we directly test whether making the late stage representations of models more similar to that of macaque IT produces new models that exhibit more robust, primate-like behavior. We conducted chronic, large-scale multi-electrode recordings across the IT cortex in six non-human primates (rhesus macaques). We then use these data to fine-tune (end-to-end) the model "IT" representations such that they are more aligned with the biological IT representations, while preserving accuracy on object recognition tasks. We generate a cohort of models with a range of IT similarity scores validated on held-out animals across two image sets with distinct statistics. Across a battery of optimization conditions, we observed a strong correlation between the models IT-likeness and alignment with human behavior, as well as an increase in its adversarial robustness. We further assessed the limitations of this approach and find that the improvements in behavioral alignment and adversarial robustness generalize across different image statistics, but not to object categories outside of those covered in our IT training set. Taken together, our results demonstrate that building models that are more aligned with the primate brain leads to more robust and human-like behavior, and call for larger neural data-sets to further augment these gains.

neuroscience↗

bioRxiv2021

Computational and biochemical analyses reveal that cofilin-2 self assembles into amyloid-like structures and promotes the aggregation of other proteinaceous species: Pathogenic relevance to myopathies

Cofilin-2 is a member of the ADF/cofilin family, expressed extensively in adult muscle cells and involved in muscle maintenance and regeneration. Phosphorylated cofilin-2 is found in pre-fibrillar aggregates formed during idiopathic dilated cardiomyopathy. A recent study shows that phosphorylated cofilin-2, under oxidative distress, forms fibrillar aggregates. However, it remains unknown if cofilin-2 holds an innate propensity to form amyloid-like structures. In the present study, we employed various computational and biochemical techniques to explore the amyloid-forming potential of cofilin-2. We report that cofilin-2 possesses aggregation-prone regions (APRs), and these APRs get exposed to the surface, become solvent-accessible, and are involved in the intermolecular interactions during dimerization, an early stage of aggregation. Furthermore, the cofilin-2 amyloids, formed under physiological conditions, are capable of cross-seeding other monomeric globular proteins and amino acids, thus promoting their aggregation. We further show that Cys-39 and Cys-80 are critical in maintaining the thermodynamic stability of cofilin-2. The destabilizing effect of oxidation at Cys-39 but not that at Cys-80 is mitigated by Ser-3 phosphorylation. Cysteine oxidation leads to partial unfolding and loss of structure, suggesting that cysteine oxidation further induces early events of cofilin-2 aggregation. Overall, our results pose a possibility that cofilin-2 amyloidogenesis might be involved in the pathophysiology of diseases, such as myopathies. We propose that the exposure of APRs to the surface could provide mechanistic insight into the higher-order aggregation and amyloidogenesis of cofilin-2. Moreover, the cross-seeding activity of cofilin-2 amyloids hints towards its involvement in the hetero-aggregation in various amyloid-linked diseases.

biochemistry↗

bioRxiv2021

A computational probe into the behavioral and neural markers of atypical facial emotion processing in autism.

Despite ample behavioral evidence of atypical facial emotion processing in individuals with autism (IwA), the neural underpinnings of such behavioral heterogeneities remain unclear. Here, I have used brain-tissue mapped artificial neural network (ANN) models of primate vision to probe candidate neural and behavior markers of atypical facial emotion recognition in IwA at an image-by-image level. Interestingly, the ANNs image-level behavioral patterns better matched the neurotypical subjects behavior than those measured in IwA. This behavioral mismatch was most remarkable when the ANN behavior was decoded from units that correspond to the primate inferior temporal (IT) cortex. ANN-IT responses also explained a significant fraction of the image-level behavioral predictivity associated with neural activity in the human amygdala -- strongly suggesting that the previously reported facial emotion intensity encodes in the human amygdala could be primarily driven by projections from the IT cortex. Furthermore, in silico experiments revealed how learning under noisy sensory representations could lead to atypical facial emotion processing that better matches the image-level behavior observed in IwA. In sum, these results identify primate IT activity as a candidate neural marker and demonstrate how ANN models of vision can be used to generate neural circuit-level hypotheses and guide future human and non-human primate studies in autism.

neuroscience↗

bioRxiv2021

Goal-Driven Recurrent Neural Network Models of the Ventral Visual Stream

The computational role of the abundant feedback connections in the ventral visual stream (VVS) is unclear, enabling humans and non-human primates to effortlessly recognize objects across a multitude of viewing conditions. Prior studies have augmented feedforward convolutional neural networks (CNNs) with recurrent connections to study their role in visual processing; however, often these recurrent networks are optimized directly on neural data or the comparative metrics used are undefined for standard feedforward networks that lack these connections. In this work, we develop task-optimized convolutional recurrent (ConvRNN) network models that more correctly mimic the timing and gross neuroanatomy of the ventral pathway. Properly chosen intermediate-depth ConvRNN circuit architectures, which incorporate mechanisms of feedforward bypassing and recurrent gating, can achieve high performance on a core recognition task, comparable to that of much deeper feedforward networks. We then develop methods that allow us to compare both CNNs and ConvRNNs to fine-grained measurements of primate categorization behavior and neural response trajectories across thousands of stimuli. We find that high performing ConvRNNs provide a better match to this data than feedforward networks of any depth, predicting the precise timings at which each stimulus is behaviorally decoded from neural activation patterns. Moreover, these ConvRNN circuits consistently produce quantitatively accurate predictions of neural dynamics from V4 and IT across the entire stimulus presentation. In fact, we find that the highest performing ConvRNNs, which best match neural and behavioral data, also achieve a strong Pareto-tradeoff between task performance and overall network size. Taken together, our results suggest the functional purpose of recurrence in the ventral pathway is to fit a high performing network in cortex, attaining computational power through temporal rather than spatial complexity.

neuroscience↗

bioRxiv2020

Fast recurrent processing via ventral prefrontal cortex is needed by the primate ventral stream for robust core visual object recognition

Distributed neural population spiking patterns in macaque inferior temporal (IT) cortex that support core visual object recognition require additional time to develop for specific ("late-solved") images suggesting the necessity of recurrent processing in these computations. Which brain circuit motifs are most responsible for computing and transmitting these putative recurrent signals to IT? To test whether the ventral prefrontal cortex (vPFC) is a critical recurrent circuit node in this system, here we pharmacologically inactivated parts of the vPFC and simultaneously measured IT population activity, while monkeys performed object discrimination tasks. Our results show that vPFC inactivation deteriorated the quality of the late-phase (>150 ms from image onset) IT population code, along with commensurate, specific behavioral deficits for "late-solved" images. Finally, silencing vPFC caused the monkeys IT activity patterns and behavior to become more like those produced by feedforward artificial neural network models of the ventral stream. Together with prior work, these results argue that fast recurrent processing through the vPFC is critical to the production of behaviorally-sufficient object representations in IT.

neuroscience↗

bioRxiv2019

A potential cortical precursor of visual word form recognition in untrained monkeys

Skilled human readers can readily recognize written letters and letter strings. This domain of visual recognition, known as orthographic processing, is foundational to human reading, but it is unclear how it is supported by neural populations in the human brain. Behavioral research has shown that non-human primates (baboons) can learn to distinguish written English words from pseudo-words (lexical decision), successfully generalize that behavior to novel strings, and exhibit behavioral error patterns that are consistent with humans. Thus, non-human primate models, while not capturing the entirety of human reading abilities, may provide a unique opportunity to investigate the neuronal mechanisms underlying orthographic processing. Here, we investigated the neuronal representation of letters and letter strings in the ventral visual stream of naive macaque monkeys, and asked to what extent these representations could support visual word recognition. We recorded the activity of hundreds of neurons at the top two levels of the ventral visual form processing pathway (V4 and IT) while monkeys passively viewed images of letters, English words, and non-word letter strings. Linear decoders were used to probe whether those neural responses could support a battery of orthographic processing tasks such as invariant letter identification and lexical decision. We found that IT-based decoders achieved baboon-level performance on these tasks, with a pattern of errors highly correlated to the previously reported primate behavior. This capacity to support orthographic processing tasks was also present in the high-layer units of state-of-the-art artificial neural network models of the ventral stream, but not in the low-layer representations of those models. Taken together, these results show that the IT cortex of untrained monkeys contains a reservoir of precursor features from which downstream brain regions could, with some supervised instruction, learn to support the visual recognition of written words. This suggests that the acquisition of reading in humans did not require a full rebuild of visual processing, but rather the recycling of a brain network evolved for other visual functions.

neuroscience↗

bioRxiv2019

Transcranial alternating current stimulation reduces BOLD adaptation and increases functional connectivity

Transcranial alternating current stimulation (tACS) is used as a non-invasive tool for cognitive enhancement and clinical applications. The physiological effects of tACS, however, are complex and poorly understood (Liu et al. 2018). Most studies of tACS focus on its ability to entrain brain oscillations (Herrmann et al. 2013), but our behavioral results in humans (Kar and Krekelberg 2014a) and extracellular recordings in nonhuman primates (Kar et al. 2017) support the view that tACS at 10 Hz additionally affects brain function by reducing sensory adaptation. Our primary goal here was to test this hypothesis using BOLD imaging in human subjects. Using a motion adaptation paradigm developed to quantify BOLD adaptation (Huk et al. 2001) and concurrent fMRI and tACS, we found that tACS significantly attenuated adaptation in the human motion area (hMT+). In addition, an exploratory analysis showed that tACS increased functional connectivity between the stimulated hMT+ and the rest of the brain, in particular the dorsal attention network. We conclude that weak 10 Hz currents applied to the scalp affect both local and global measures of brain activity.\n\nNew and NoteworthyConcurrent transcranial alternating current stimulation (tACS) and fMRI show that tACS affects the human brain by attenuating adaptation and increasing functional connectivity. This work is important for our basic understanding of what tACS does, but also for therapeutic applications, which need insight into the full range of ways in which tACS affects the brain.

neuroscience↗

bioRxiv2018

Neural Population Control via Deep Image Synthesis

Particular deep artificial neural networks (ANNs) are todays most accurate models of the primate brains ventral visual stream. Here we report that, using a targeted ANN-driven image synthesis method, new luminous power patterns (i.e. images) can be applied to the primate retinae to predictably push the spiking activity of targeted V4 neural sites beyond naturally occurring levels. More importantly, this method, while not yet perfect, already achieves unprecedented independent control of the activity state of entire populations of V4 neural sites, even those with overlapping receptive fields. These results show how the knowledge embedded in todays ANN models might be used to non-invasively set desired internal brain states at neuron-level resolution, and suggest that more accurate ANN models would produce even more accurate control.

neuroscience↗