Biology subjects

Kording, K. P.

Publications and source records attributed to Kording, K. P..

At least 19 recordsLinked to original sources

bioRxiv2026

Widespread use of invalid statistical tests in biomedical machine learning

Cross-validation is routinely used to compare performance in biomedical artificial intelligence. Standard tests ignore correlation across cross-validation folds, inflating false-positive rates. In a PRISMA-guided meta-analysis of 184 studies (impact factor [≥] 15) across 30 biomedical fields, 97% use invalid tests. Among studies with abstract-level claims supported by invalid tests, 59% rely on a spurious comparison: one that loses significance after correcting for this correlation. On average, regaining significance requires a 43% larger effect. Extending these findings to the broader literature, we estimate that spurious comparisons occur in nearly two in three studies and support abstract-level claims in nearly one in three studies. Simulations confirm false-positive rates paradoxically approach 100% when cross-validation is repeated to improve stability. We introduce SHARP, a redesign of cross-validation, which best balances false-positive control and power among 13 benchmarked tests. These results reveal widespread fragility in biomedical artificial intelligence and offer a practical route to valid comparisons.

bioinformatics↗

bioRxiv2026

Physical contact reveals a hidden layer of cortical architecture

Neurons interact at synapses, but they also communicate through physical contact and proximity, including diffusion, glia-mediated interactions, and ephaptic coupling. Standard connectomes map synapses, but cannot capture the full set of cell-cell contacts that can support these pathways. Here we extract contactomes from two large mouse visual cortex volumes at nanoscale resolution and quantify every cell-cell contact, the shared surface area of each contact, and the relationship between contact and synaptic connectivity. We find that contactomes are 5 - 10x denser than synaptic graphs, revealing that neurons physically contact a much larger set of potential neighbors than they synaptically connect to. We further find that most nearby potential neighbors are already in physical contact, indicating that local structural change would add few new candidate synaptic partners. Finally, we find that astrocytes form a single large syncytium-like network that spans the tissue and directly contacts nearly all neurons, and that glial processes lie within a micron or two of almost every synapse, indicating that synapses reside within a pervasive glia-shaped microenvironment. Together, these results show that physical contact forms a distinct layer of brain architecture that extends far beyond the synaptic connectome.

neuroscience↗

bioRxiv2026

Mid-superior temporal sulcus encodes spatial context and behavioral state in freely moving macaques

Primate neuroscience has traditionally studied the brain under highly constrained conditions, limiting our understanding of neural function during real-world behavior. The mid-superior temporal sulcus (mSTS) is implicated in social perception, but its role during unconstrained behavior has not been tested. Here we performed wireless depth-electrode recordings from both banks of mSTS in macaques freely exploring a large three-dimensional arena, combined with 3D pose tracking and behavioral segmentation. Neural encoding models revealed mSTS firing rates were jointly modulated by spatial position, body kinematics, and geometric visual proxies, preferentially encoded in allocentric coordinates but with a shift toward body-centric encoding during vertical exploration. Neural populations carried decodable information about discrete behavioral syllables, with broad temporal generalization and neural similarity that tracked the sequential structure of behavior. Population manifold analysis revealed that the same behavior occupied different regions of population space at different spatial locations, and population dynamics showed structured organization around behavioral transitions. Together, these results suggest that mSTS populations carry joint information about spatial context and behavioral state during natural behavior.

neuroscience↗

bioRxiv2025

Resting-State fMRI and the Risk of Overinterpretation: Noise, Mechanisms, and a Missing Rosetta Stone

Resting-state fMRI has generated influential insights into large-scale brain organization and contributed to clinically relevant applications, largely through correlation-based measures of cross-regional association in BOLD responses. At the same time, interpreting these statistical associations as reflecting underlying neural interactions requires careful consideration of fundamental methodological constraints. Here, we distinguish two fundamental but often conflated limitations. The first is measurement distortion: the fMRI signal is an indirect and heterogeneous measurement of neural activity, arising from neurovascular coupling, physiology, and measurement-related processes, which can introduce systematic and incompletely characterized biases into estimated correlations. The second is causal non-identifiability: even if correlations perfectly reflected neural synchrony, the resulting correlation structure would not uniquely determine the underlying neural interactions. Using causal reasoning, simulations, and analytic arguments, we distinguish these limitations and examine their consequences for interpretation. Although both apply broadly to fMRI, their implications are particularly important in resting-state analyses, where correlation structure is the primary object of inference in the absence of experimental perturbation. We show that measurement-related biases can distort estimated correlations (e.g., attenuating or in-flating associations and affecting group comparisons) and can produce reproducible patterns that do not necessarily reflect underlying neural relationships, highlighting that statistical reliability does not guarantee biological validity. We further show that graph-theoretic, geometric, and other higher-order representations derived from these correlations do not, by themselves, justify mechanistic interpretation. We argue not against the utility of resting-state fMRI, but for greater precision in interpretation. Our conclusions concern the interpretation of correlation-based analyses rather than their methodological utility. Distinguishing measurement distortion from causal non-identifiability clarifies the inferential boundaries separating descriptive association, predictive utility, and mechanistic interpretation.

neuroscience↗

bioRxiv2025

PrimateFace: A Machine Learning Resource for Automated Face Analysis in Human and Non-human Primates

Machine learning has revolutionized human face analysis, but equivalent tools for non-human primates remain limited and species-specific, hindering progress in neuroscience, anthropology, and conservation. Here, we present PrimateFace, a comprehensive, cross-species platform for primate facial analysis comprising a systematically curated dataset of 260,000+ images spanning over 60 genera, including a genus-balanced subset of 60,000 images, annotated with bounding boxes and facial landmark configurations. Face detection and facial landmark estimation models trained on PrimateFace achieve high cross-species performance, from tarsiers to gorillas, achieving performance comparable to baseline models trained exclusively on human data (0.34 vs. 0.39 mAP for face detection; 0.061 vs. 0.053 normalized landmark error), demonstrating the generalization benefits of cross-species training. PrimateFace enables diverse downstream applications including individual recognition, gaze analysis, and automated extraction of stereotyped (e.g., lip-smacking) and subtle (e.g., soft left turn) facial movements. PrimateFace provides a standardized platform for facial phenotyping across the primate order, empowering data-driven studies that advance the health and well-being of human and non-human primates. All models, notebooks, and data can be found at github.com/KordingLab/PrimateFace.

animal behavior and cognition↗

bioRxiv2025

Humans can learn bimodal priors in complex sensorimotor behaviour

Extensive research suggests that humans integrate sensory information and prior expectations in a Bayesian manner to reduce uncertainty in perception and action. However, while Bayesian integration provides a powerful explanatory framework, the question remains as to what extent it explains human behaviour in naturalistic situations, including more complex movements and distributions. Here, we examine whether humans can learn bimodal priors in a complex sensorimotor task: returning tennis serves. Participants returned serves in an immersive virtual reality setup with realistic movements and spatiotemporal task demands matching those in real tennis. The location of the opponents serves followed a bimodal distribution. We manipulated visual uncertainty through three levels of ball speeds: slow, moderate, and fast. After extensive exposure to the opponents serves, participants movements were biased by the bimodal prior distribution. As predicted by Bayesian theory, the magnitude of the bias depends on visual uncertainty. Additionally, our data indicate that participants movements in this complex task were not only biased by prior expectations but also by biomechanical constraints and associated motor costs. Intriguingly, an explicit knowledge test after the experiment revealed that, despite incorporating prior knowledge of the opponents serve distribution into their behaviour, participants were not explicitly aware of the pattern. Our results show that humans can implicitly learn and utilise bimodal priors in complex sensorimotor behaviour.

neuroscience↗

bioRxiv2024

Automated segmentation of synchrotron-scanned fossils

Computed tomography has revolutionised the study of the internal three-dimensional structure of fossils. Historically, fossils typically spent years in preparation to be freed from the enclosing rock. Now, X-ray and synchrotron tomography reveal structure that is otherwise invisible and data acquisition can be fast. However, manual segmentation of these 3D volumes can still take months to years. This is especially challenging for resource-poor teams, as scanning may be free, but the computing power and (AI-assisted) segmentation software required to handle the resulting large data sets are complex to use and expensive. Here we present a free, browser-based segmentation tool that reduces computational overhead by splitting volumes into small chunks, allowing processing on low-memory, inexpensive hardware. Our tool also speeds up collaborative ground-truth generation and 3D visualization, all in-browser. We developed and evaluated our pipeline on various open-data scans of differing contrast, resolution, textural complexity, and size. Our tool successfully isolated the Thrinaxodon and Broomistega pair from an Early Triassic burrow. It isolated cranial bones from the Cretaceous acipenseriform Parapsephurus willybemisi on both 45.53 {micro}m and 13.67 {micro}m resolution scanning data. We also isolated bones of the Middle Triassic sauropterygian Nothosaurus and a challenging scan of a squamate embryo inside an egg dating back to the Early Cretaceous. Our tool reliably reproduces expert-supervised segmentation at a fraction of the time and cost, offering greater accessibility than existing tools. Beyond the online tool, all our code is open source, enabling contributions from the palaeontology community to further this emerging machine learning ecosystem.

paleontology↗

bioRxiv2024

Individual-specific strategies inform category learning

Categorization is an essential task for sensory perception. Individuals learn category labels using a variety of strategies to ensure that sensory signals, such as sounds or images, can be assigned to proper categories. Categories are often learned on the basis of extreme examples, and the boundary between categories can differ among individuals. The trajectories for learning also differ among individuals, as different individuals rely on different strategies, such as repeating or alternating choices. However, little is understood about the relationship between individual learning trajectories and learned categorization. To study this relationship, we trained mice to categorize auditory stimuli into two categories using a two-alternative forced choice task. Because the mice took several weeks to learn the task, we were able to quantify the time course of individual strategies and how they relate to how mice categorize stimuli around the categorization boundary. Different mice exhibited different trajectories while learning the task. Mice displayed preferences for a specific category, manifested by a choice bias in their responses, but this bias drifted with learning. We found that this drift in choice bias correlated with variability in the category boundary for sounds with ambiguous category membership. Next, we asked how stimulus-independent, individual-specific strategies informed learning. We found that the tendency to repeat choices, which is a form of perseveration, contributed to long-term learning. These results indicate that long-term trends in individual strategies during category learning affect learned category boundaries.

neuroscience↗

bioRxiv2024

Modeling Attention and Binding in the Brain through Bidirectional Recurrent Gating

Attention is a key component of the visual system, essential for perception, learning, and memory. Attention can also be seen as a solution to the binding problem: concurrent attention to all parts of an entity allows separating it from the rest. However, the rich models of attention in computational neuroscience are generally not scaled to real-world problems and there are thus many behavioral and neural phenomena that current models cannot explain. Here, we propose a bidirectional recurrent model of attention that is inspired by the emerging understanding of biological object-based attention and modern neural networks for image segmentation. It conceptualizes recurrent connections as a multi-stage internal gating process where bottom-up connections transmit features, while top-down and lateral connections transmit attentional gating signals. Our model can recognize and segment simple stimuli such as digits as well as objects in natural images and is able to be prompted with object labels, attributes or locations. It can learn to perform a range of behavioral findings, such as object binding, selective attention, inhibition of return, and visual search. It also replicates a variety of neural findings, including increased activity for attended objects, features, and locations, attention-invariant tuning, and relatively late onset attention. Most importantly, our proposed model unifies decades of cognitive and neurophysiological findings of visual attention into a single principled architecture. Our results highlight that the ability to selectively and dynamically focus on specific parts of stimulus streams can help artificial neural networks to better generalize and align with human brains.

neuroscience↗

bioRxiv2024

Conclusions about Neural Network to Brain Alignment are Profoundly Impacted by the Similarity Measure

Deep neural networks are widely used to model biological perception and behavior, making the similarities and differences between artificial and biological systems consequential. If a principle (e.g. self-supervised learning) produces a model resembling biology, this is taken as evidence the same principle shapes the biological system. But what it means for an artificial system to be similar to a biological one is complex. A popular approach compares representations of identical stimuli using a similarity measure like Representational Similarity Analysis. Yet scientific questions rarely specify which measure is appropriate, raising a key question: do conclusions depend on this choice? Focusing on vision, we show that measure choice influences both hierarchical correspondence between systems and the ranking of which artificial systems are most biological. Reanalyzing prior studies, we find the choice hugely consequential: models best under one measure are often worst under another, and prior conclusions can flip or dissolve. Different metrics capture fundamentally different aspects of similarity, warranting healthy skepticism toward such comparisons.

neuroscience↗

bioRxiv2023

Neural signatures of natural behavior in socializing macaques

Our understanding of the neurobiology of primate behavior largely derives from artificial tasks in highly-controlled laboratory settings, overlooking most natural behaviors primate brains evolved to produce1-3. In particular, how primates navigate the multidimensional social relationships that structure daily life4 and shape survival and reproductive success5 remains largely unexplored at the single neuron level. Here, we combine ethological analysis with new wireless recording technologies to uncover neural signatures of natural behavior in unrestrained, socially interacting pairs of rhesus macaques. Single neuron and population activity in prefrontal and temporal cortex unveiled robust encoding of 24 species-typical behaviors, which was strongly modulated by the presence and identity of surrounding monkeys. Male-female partners demonstrated near-perfect reciprocity in grooming, a key behavioral mechanism supporting friendships and alliances6, and neural activity maintained a running account of these social investments. When confronted with an aggressive intruder, behavioral and neural population responses reflected empathy and were buffered by the presence of a partner. By employing an ethological approach to the study of primate neurobiology, we reveal a highly-distributed neurophysiological ledger of social dynamics, a potential computational foundation supporting communal life in primate societies, including our own.

neuroscience↗

bioRxiv2023

When Neural Activity Fails to Reveal Causal Contributions

Neuroscientists rely on distributed spatio-temporal patterns of neural activity to understand how neural units contribute to cognitive functions and behavior. However, the extent to which neural activity reliably indicates a units causal contribution to the behavior is not well understood. To address this issue, we provide a systematic multi-site perturbation framework that captures time-varying causal contributions of elements to a collectively produced outcome. Applying our framework to intuitive toy examples and artificial neuronal networks revealed that recorded activity patterns of neural elements may not be generally informative of their causal contribution due to activity transformations within a network. Overall, our findings emphasize the limitations of inferring causal mechanisms from neural activities and offer a rigorous lesioning framework for elucidating causal neural contributions.

neuroscience↗

bioRxiv2022

The interplay of uncertainty, relevance and learning influences auditory categorization

Auditory perception requires categorizing sound sequences, such as speech or music, into classes, such as syllables or notes. Auditory categorization depends not only on the acoustic waveform, but also on variability and uncertainty in how the listener perceives the sound - including sensory and stimulus uncertainty, the listeners estimated relevance of the particular sound to the task, and their ability to learn the past statistics of the acoustic environment. Whereas these factors have been studied in isolation, whether and how these factors interact to shape categorization remains unknown. Here, we measured human participants performance on a multi-tone categorization task and modeled each participants behavior using a Bayesian framework. Task-relevant tones contributed more to category choice than task-irrelevant tones, confirming that participants combined information about sensory features with task relevance. Conversely, participants poor estimates of task-relevant tones or high-sensory uncertainty adversely impacted category choice. Learning the statistics of sound category over both short and long timescales also affected decisions, biasing the decisions toward the overrepresented category. The magnitude of this effect correlated inversely with participants relevance estimates. Our results demonstrate that individual participants idiosyncratically weigh sensory uncertainty, task relevance, and statistics over both short and long timescales, providing a novel understanding of and a computational framework for how sensory decisions are made under several simultaneous behavioral demands.

neuroscience↗

bioRxiv2022

Cell lineage predicts neural connectivity beyond cell type

During nervous system development, repeated cell divisions of the zygote give rise to a "family tree" of neurons related by mitotic lineage. The developmental process also gives rise to neuronal connections, and neurons phenotypically converge to different cell types. Neural connections are steered by cell type, but they may be driven in part by lineage: We do not know if lineage matters for the developmental neurogenesis process. We thus asked if mitotic lineage predicts neural connections beyond cell type alone. Using three C. elegans datasets, we fit models for edge prediction tasks: predicting synaptic targets, and predicting synaptic sources. Adding the mitotic lineage improved these predictions. Our results suggest that the family tree matters for the connections made by neurons, and that developmental lineage is a variable that should be considered more deeply in models of connectome assembly.

neuroscience↗

bioRxiv2022

Bayesball: Bayesian Integration in Professional Baseball Batters

Pitchers in baseball throw the ball with such high velocity and varying movement that batters only have a few hundred milliseconds to estimate whether to swing and how high to swing. Slight deviations in the contact point on the ball can result in weakly hit balls that do not result in opportunities for batters to score. Even before the pitcher releases the ball, the batter has some belief (an estimated distribution-a prior), of where the ball may land in the strike zone. Batters will update this prior belief with information from observing the pitch (the likelihood) to calculate their final estimate (the posterior). These models of behavior, called Bayesian models within movement science, predict that players will estimate a final ball position by combining prior information with observation of the pitch in a way that weights each information source relative to the uncertainty. Here we test this model using data from more than a million samples from professional baseball. Moreover, as predicted by a Bayesian model, we show that a batters estimate of where to swing is biased towards the prior when prior information is available ( pitch tipping), and biased towards the likelihood in the case of pitches with high movement uncertainty. These results demonstrate that Bayesian ideas are relevant well beyond laboratory experiments and matter for real-world movements.

neuroscience↗

bioRxiv2022

Efficient neural codes naturally emerge through gradient descent learning

Animal sensory systems are more sensitive to common features in the environment than uncommon features. For example, small deviations from the more frequently encountered horizontal orientations can be more easily detected than small deviations from the less frequent diagonal ones. Here we find that artificial neural networks trained to recognize objects also have patterns of sensitivity that match the statistics of features in images. To interpret these findings, we show mathematically that learning with gradient descent in deep neural networks preferentially creates representations that are more sensitive to common features, a hallmark of efficient coding. This result suggests that efficient coding naturally emerges from gradient-like learning on natural stimuli.

neuroscience↗

bioRxiv2021

Recording Temporal Signals with Minutes Resolution Using Enzymatic DNA Synthesis

Employing DNA as a high-density data storage medium has paved the way for next-generation digital storage and biosensing technologies. However, the multipart architecture of current DNA-based recording techniques renders them inherently slow and incapable of recording fluctuating signals with sub-hour frequencies. To address this limitation, we developed a simplified system employing a single enzyme, terminal deoxynucleotidyl transferase (TdT), to transduce environmental signals into DNA. TdT adds nucleotides to the 3 ends of single-stranded DNA (ssDNA) in a template-independent manner, selecting bases according to inherent preferences and environmental conditions. By characterizing TdT nucleotide selectivity under different conditions, we show that TdT can encode various physiologically relevant signals like Co2+, Ca2+, Zn2+ concentrations and temperature changes in vitro. Further, by considering the average rate of nucleotide incorporation, we show that the resulting ssDNA functions as a molecular ticker tape. With this method we accurately encode a temporal record of fluctuations in Co2+ concentration to within 1 minute over a 60-minute period. Finally, we engineer TdT to allosterically turn off in the presence of physiologically relevant concentration of calcium. We use this engineered TdT in concert with a reference TdT to develop a two-polymerase system capable of recording a single step change in Ca2+ signal to within 1 minute over a 60-minute period. This work expands the repertoire of DNA-based recording techniques by developing a novel DNA synthesis-based system that can record temporal environmental signals into DNA with minutes resolution.

synthetic biology↗

bioRxiv2021

Energy expenditure does not explain step length-width choices during walking

Healthy young adults have a most preferred walking speed, step length, and step width that are close to energetically optimal. However, people can choose to walk with a multitude of different step lengths and widths, which can vary in both energy expenditure and preference. Here we further investigate step length-width preferences and their relationship to energy expenditure. In line with a growing body of research, we hypothesized that peoples preferred stepping patterns would not be fully explained by metabolic energy expenditure. To test this hypothesis we used a two-alternative forced-choice paradigm. Fifteen participants walked on an oversized treadmill. Each trial participants experienced two stepping patterns and then chose the pattern they preferred. Over time, we adapted the choices such that there was 50% chance of choosing one pattern over another (equally preferred). If peoples preferences are based solely on metabolic energy expenditure, then these equally preferred stepping patterns should have equal energy expenditure. We found that energy expenditure differed across equally preferred step length-width patterns (p < 0.001). On average, longer steps with higher energy expenditures were preferred over shorter and wider steps with lower energy expenditures (p < 0.001). We also asked participants to rank a set of shorter, wider, and longer steps from most preferred to least preferred, and from most energy expended to least energy expended. Only 7/15 participants had the same rankings for their preferences and perceived energy expenditure. Our results suggest that energy expenditure is not the only factor influencing a persons conscious gait choices.

neuroscience↗