Publications

The Poisson Generalized Linear Model (GLM) is a foundational tool for analyzing neural spike train data. However, standard implementations rely on discretizing spike times into binned count data, limiting temporal resolution and scalability. Here, we develop Monte Carlo (MC) methods and polynomial approximations (PA) to the continuous-time analog of these models, and show them to be advantageous over their discrete-time counterparts. Further, we propose using a set of exponentially scaled Laguerre polynomials as an orthogonal temporal basis, which improves filter identification and yields closed-form integral solutions under the polynomial approximation. Applied to both synthetic and real spike-time data from rodent hippocampus, our methods demonstrate superior accuracy and scalability compared to traditional binned GLMs, enabling functional connectivity inference in large-scale neural recordings that are temporally precise on the order of synaptic dynamical timescales and in agreement with known anatomical properties of hippocampal subregions. We provide open-source implementations of both MC and PA estimators, optimized for GPU acceleration, to facilitate adoption in the neuroscience community.

Large Language Models (LLMs) are increasingly being used across applications, ranging from content generation to decision-making, raising concerns about biases embedded in them. While biases related to gender, race, and culture have been studied extensively, understanding age-bias and stereotypes of aging in LLMs remain underexplored. This study analyzes LLM-generated responses to prompts related to aging, revealing stereotypical biases about aging pertaining to technology proficiency, cognitive and physical decline, and job roles. We noted that even responses without explicit age bias also had mentions of ageist stereotypes. We discuss considerations for involving older adults’ perspectives through human-in-the-loop methodologies yet exercising caution about aspects like internalized ageism.

Generative diffusion models learn probability densities over diverse image datasets by estimating the score with a neural network trained to remove noise. Despite their remarkable success in generating high-quality images, the internal mechanisms of the underlying score networks are not well understood. Here, we examine the image representation that arises from score estimation in a {fully-convolutional unconditional UNet}. We show that the middle block of the UNet decomposes individual images into sparse subsets of active channels, and that the vector of spatial averages of these channels can provide a nonlinear representation of the underlying clean images. Euclidean distances in this representation space are semantically meaningful, even though no conditioning information is provided during training. We develop a novel algorithm for stochastic reconstruction of images conditioned on this representation: The synthesis using the unconditional model is "self-guided" by the representation extracted from that very same model. For a given representation, the common patterns in the set of reconstructed samples reveal the features captured in the middle block of the UNet. Together, these results show, for the first time, that a measure of semantic similarity emerges, unsupervised, solely from the denoising objective.

We introduce the Information-Estimation Metric (IEM), a novel form of distance function derived from an underlying continuous probability density over a domain of signals. The IEM is rooted in a fundamental relationship between information theory and estimation theory, which links the log-probability of a signal with the errors of an optimal denoiser, applied to noisy observations of the signal. In particular, the IEM between a pair of signals is obtained by comparing their denoising error vectors over a range of noise amplitudes. Geometrically, this amounts to comparing the score vector fields of the blurred density around the signals over a range of blur levels. We prove that the IEM is a valid global metric and derive a closed-form expression for its local second-order approximation, which yields a Riemannian metric. For Gaussian-distributed signals, the IEM coincides with the Mahalanobis distance. But for more complex distributions, it adapts, both locally and globally, to the geometry of the distribution. In practice, the IEM can be computed using a learned denoiser (analogous to generative diffusion models) and solving a one-dimensional integral. To demonstrate the value of our framework, we learn an IEM on the ImageNet database. Experiments show that this IEM is competitive with or outperforms state-of-the-art supervised image quality metrics in predicting human perceptual judgments.

Motivation Focused Ion Beam Scanning Electron Microscopy (FIB-SEM) is an advanced Volume Electron Microscopy technology with growing applications, featuring thinner sectioning compared to other Volume Electron Microscopes. Such axial resolution is crucial for accurate segmentation and reconstruction of fine structures in biological tissues. However, in reality, the milling thickness is not always uniform across the sample surface, resulting in the axial plane looking distorted. Existing image processing approaches often: (i) assume constant section thickness; (ii) consist of multiple separate processing steps (i.e., not in an end-to-end fashion); (iii) require ground truth images for modeling, which may entail significant labor and be unsuitable for rapid analysis.

Results We develop a deep learning method to correct non-uniform milling artifacts observed in FIB-SEM images. The proposed method is an image-to-image translation technique that can mitigate image distortions in an unsupervised manner. It conducts cross-plane learning within 3D image volumes without any ground truth annotations. We demonstrate the efficacy of our method on a real-world micro-wasp dataset, showcasing significantly improved image quality after correction with qualitative and quantitative analysis.

IgA nephropathy (IgAN), focal segmental glomerulosclerosis (FSGS), membranous nephropathy (MN), and minimal change disease (MCD) account for the majority of idiopathic glomerulopathies (GN). However, there are no powered transcriptomic datasets coupled to genetic data to investigate the genetic mechanisms underlying gene regulation in the context of GN.

The global dimensionality of a neural representation manifold provides rich insight into the computational process underlying both artificial and biological neural networks. However, all existing measures of global dimensionality are sensitive to the number of samples, i.e., the number of rows and columns of the sample matrix. We show that, in particular, the participation ratio of eigenvalues, a popular measure of global dimensionality, is highly biased with small sample sizes, and propose a bias-corrected estimator that is more accurate with finite samples and with noise. On synthetic data examples, we demonstrate that our estimator can recover the true known dimensionality. We apply our estimator to neural brain recordings, including calcium imaging, electrophysiological recordings, and fMRI data, and to the neural activations in a large language model and show our estimator is invariant to the sample size. Finally, our estimators can additionally be used to measure the local dimensionalities of curved neural manifolds by weighting the finite samples appropriately.

The human auditory cortex is topographically organized. Neurons with similar response properties are spatially clustered, forming smooth maps for acoustic features such as frequency in early auditory areas, and modular regions selective for music and speech in higher-order cortex. Yet, evaluations for current computational models of auditory perception do not measure whether such topographic structure is present in a candidate model. Here, we show that cortical topography is not present in the previous best-performing models at predicting human auditory fMRI responses. To encourage the emergence of topographic organization, we adapt a cortical wiring-constraint loss originally designed for visual perception. The new class of topographic auditory models, TopoAudio, are trained to classify speech, and environmental sounds from cochleagram inputs, with an added constraint that nearby units on a 2D cortical sheet develop similar tuning. Despite these additional constraints, TopoAudio achieves high accuracy on benchmark tasks comparable to the unconstrained non-topographic baseline models. Further, TopoAudio predicts the fMRI responses in the brain as well as standard models, but unlike standard models, TopoAudio develops smooth, topographic maps for tonotopy and amplitude modulation (common properties of early auditory representation, as well as clustered response modules for music and speech (higher-order selectivity observed in the human auditory cortex). TopoAudio is the first end-to-end biologically grounded auditory model to exhibit emergent topography, and our results emphasize that a wiring-length constraint can serve as a general-purpose regularization tool to achieve biologically aligned representations.

While the thalamus is known to relay and modulate sensory signals to the cortex, whether it also participates in active computation and intrinsic signal generation remains unresolved. The anterodorsal nucleus of the thalamus broadcasts the head-direction (HD) signal, which is generated in the brainstem, particularly in the upstream lateral mammillary nucleus, and thalamic HD cells remain coordinated even during sleep. Here, by recording and manipulating neuronal activity along the mammillary–thalamic–cortical pathway, we show that coherence among thalamic HD cells persists even when their upstream inputs are decorrelated, particularly during non-Rapid Eye Movement sleep. These findings suggest that thalamic circuits are sufficient to generate and maintain coherent population dynamics in the absence of structured input.

The segregation of the epiblast (EPI) and primitive endoderm (PE) cell types in the preimplantation mouse embryo is not only a crucial decision that sets aside the precursors of the embryo proper from extraembryonic cells, respectively, but also has served as a central model to study a key concept in mammalian development: how much of developmental patterning is predetermined vs. stochastically emergent. Here, we address this question by quantitative live imaging of multiple endogenously tagged transcription factors key to this fate decision and trace their dynamics at a single-cell resolution through the formation of EPI and PE cell fates. Strikingly, we reveal an initial symmetry breaking event, the formation of a primary EPI cell lineage, and show that this is linked to the dynamics of the prior inner cell mass/trophectoderm fate decision through the expression of SOX2. This primary EPI lineage, through fibroblast growth factor (FGF) signaling, induces an increase in the transcription factor GATA6 in other inner cell mass cells, setting them on the course toward PE differentiation. Interestingly, this trajectory can switch during a defined developmental window, leading to the emergence of secondary EPI cells. Finally, we show that early expression levels of NANOG, which are seemingly stochastic, can bias whether a cell’s trajectory switches to secondary EPI or continues as PE. Our data give unique insight into how fate patterning is initiated and propagated during unperturbed embryonic development through the interplay of lineage-history-biased and stochastic cell-intrinsic molecular features, unifying previous models of EPI/PE segregation.