About

Shreya Saxena is an Assistant Professor in the Biomedical Engineering Department at Yale University and a core member of the Center for Neurocomputation and Machine Intelligence at the Wu Tsai Institute. Her research broadly focuses on the neural control of complex, coordinated behavior, with an emphasis on understanding the relationship between neural activity and behavior through constraints-based modeling approaches that incorporate anatomy and physiology. Saxena's work aims to improve the inference of quantitative dynamical models for cognition and motor control, addressing challenges in large-scale neural and behavioral data analysis. Her academic background includes a Ph.D. from the Massachusetts Institute of Technology in Electrical Engineering and Computer Science, where she studied the closed-loop control of fast movements from a control theory perspective. She also holds an M.S. in Biomedical Engineering from Johns Hopkins University and a B.S. in Mechanical Engineering from the Swiss Federal Institute of Technology (EPFL). Prior to her current role, she was an Assistant Professor at the University of Florida's Department of Electrical and Computer Engineering and a Swiss National Science Foundation Postdoctoral Fellow at Columbia University’s Zuckerman Mind Brain Behavior Institute. Saxena has been recognized as a Rising Star in both Electrical Engineering and Biomedical Engineering and was awarded a Sloan Research Fellowship in 2025.

Research topics

• Artificial Intelligence
• Computer Science
• Physics
• Neuroscience
• Mathematics
• Medicine
• Engineering
• Chemistry
• Biology
• Psychology

Selected publications

• FLUX: Geometry-Aware Longitudinal Flow Matching with Mixture of Experts

  arXiv (Cornell University) · 2026-05-09

  preprintOpen accessSenior author

  Many biological systems evolve through continuous local dynamics while switching between latent regimes defined by learning, stimulus context, internal state, or developmental stage. These processes are often observed only as unpaired longitudinal snapshots: the same cells, neurons, or animals are not tracked as matched trajectories, even though population states are sampled across successive stages. This creates two coupled challenges. First, trajectories must respect curved low-dimensional manifolds embedded in high-dimensional biological measurements. Second, the model must identify when the transport mechanism itself changes. We introduce FLUX (FLow matching for Unpaired longitudinal data with miXture-of-experts), a geometry-aware longitudinal flow-matching framework for joint transport modeling and unsupervised regime discovery. FLUX learns a data-dependent metric from pooled labeled and unlabeled observations, uses that metric to construct geometry-aware conditional paths between adjacent marginals, and decomposes the resulting velocity field into sparse expert vector fields selected by a Straight-Through Gumbel-Softmax router. Across manifold controls, a regime-switching Lorenz system, widefield cortical calcium imaging during associative learning, and embryoid body single-cell differentiation, FLUX reconstructs longitudinal transport while recovering interpretable regime structure. Ablations show that mixture-of-experts routing alone is insufficient: FLUX without geometric learning can fit local transport but fails or weakens regime discovery when regimes are encoded in local dynamics. These results suggest that geometry-aware velocity decomposition provides a general strategy for discovering latent biological state transitions from unpaired longitudinal snapshots.

  PublisherDOI

• Learning Mixtures of Linear Dynamical Systems via Hybrid Tensor-EM Method.

  PubMed · 2026-02-27

  articleSenior author

  Mixtures of linear dynamical systems (MoLDS) provide a path to model time-series data that exhibit diverse temporal dynamics across trajectories. However, its application remains challenging in complex and noisy settings, limiting its effectiveness for neural data analysis. Tensor-based moment methods can provide global identifiability guarantees for MoLDS, but their performance degrades under noise and complexity. Commonly used expectation-maximization (EM) methods offer flexibility in fitting latent models but are highly sensitive to initialization and prone to poor local minima. Here, we propose a tensor-based method that provides identifiability guarantees for learning MoLDS, which is followed by EM updates to combine the strengths of both approaches. The novelty in our approach lies in the construction of moment tensors using the input-output data to recover globally consistent estimates of mixture weights and system parameters. These estimates can then be refined through a Kalman EM algorithm, with closed-form updates for all LDS parameters. We validate our framework on synthetic benchmarks and real-world datasets. On synthetic data, the proposed Tensor-EM method achieves more reliable recovery and improved robustness compared to either pure tensor or randomly initialized EM methods. We then analyze neural recordings from the primate somatosensory cortex while a non-human primate performs reaches in different directions. Our method successfully models and clusters different conditions as separate subsystems, consistent with supervised single-LDS fits for each condition. Finally, we apply this approach to another neural dataset where monkeys perform a sequential reaching task. These results demonstrate that MoLDS provides an effective framework for modeling complex neural data, and that Tensor-EM is a reliable approach to MoLDS learning for these applications.

  Publisher

• FLUX: Geometry-Aware Longitudinal Flow Matching with Mixture of Experts

  ArXiv.org · 2026-05-09

  articleOpen accessSenior author

  Many biological systems evolve through continuous local dynamics while switching between latent regimes defined by learning, stimulus context, internal state, or developmental stage. These processes are often observed only as unpaired longitudinal snapshots: the same cells, neurons, or animals are not tracked as matched trajectories, even though population states are sampled across successive stages. This creates two coupled challenges. First, trajectories must respect curved low-dimensional manifolds embedded in high-dimensional biological measurements. Second, the model must identify when the transport mechanism itself changes. We introduce FLUX (FLow matching for Unpaired longitudinal data with miXture-of-experts), a geometry-aware longitudinal flow-matching framework for joint transport modeling and unsupervised regime discovery. FLUX learns a data-dependent metric from pooled labeled and unlabeled observations, uses that metric to construct geometry-aware conditional paths between adjacent marginals, and decomposes the resulting velocity field into sparse expert vector fields selected by a Straight-Through Gumbel-Softmax router. Across manifold controls, a regime-switching Lorenz system, widefield cortical calcium imaging during associative learning, and embryoid body single-cell differentiation, FLUX reconstructs longitudinal transport while recovering interpretable regime structure. Ablations show that mixture-of-experts routing alone is insufficient: FLUX without geometric learning can fit local transport but fails or weakens regime discovery when regimes are encoded in local dynamics. These results suggest that geometry-aware velocity decomposition provides a general strategy for discovering latent biological state transitions from unpaired longitudinal snapshots.

  PublisherOA PDF

• Integrative neurocybernetic modeling in the era of large-scale neuroscience

  arXiv (Cornell University) · 2026-04-26

  preprintOpen access

  Large-scale neuroscience is generating rich datasets across animals, brain areas and behavioral contexts, yet our modeling efforts remains fragmented across isolated experiments. We argue that understanding behavior requires integrative neurocybernetic models: understandable dynamical models that capture the closed-loop coupling of brain, body and environment, treat the brain as a controller pursuing latent objectives, represent structured variation across scales, and scale to heterogeneous datasets. Such models shift the goal from predicting neural recordings in isolation to inferring the organizing principles that govern neural and behavioral dynamics. We outline a practical route toward this goal by combining nonlinear state-space models and meta-dynamical extensions with scalable inference, knowledge distillation, mixed open- and closed-loop training, and connectomics-informed architectures. By pooling complementary constraints from recordings, behavior, perturbations and anatomy, integrative neurocybernetic models can provide statistical amplification, few-shot generalization, and mechanistic insight into shared dynamical structure, individual variation, and the control objectives that govern behavior. This agenda offers a model-centric path from fragmented data to a mechanistic science of how brains produce behavior.

  PublisherDOI

• Integrative neurocybernetic modeling in the era of large-scale neuroscience

  ArXiv.org · 2026-04-26

  articleOpen access

  Large-scale neuroscience is generating rich datasets across animals, brain areas and behavioral contexts, yet our modeling efforts remains fragmented across isolated experiments. We argue that understanding behavior requires integrative neurocybernetic models: understandable dynamical models that capture the closed-loop coupling of brain, body and environment, treat the brain as a controller pursuing latent objectives, represent structured variation across scales, and scale to heterogeneous datasets. Such models shift the goal from predicting neural recordings in isolation to inferring the organizing principles that govern neural and behavioral dynamics. We outline a practical route toward this goal by combining nonlinear state-space models and meta-dynamical extensions with scalable inference, knowledge distillation, mixed open- and closed-loop training, and connectomics-informed architectures. By pooling complementary constraints from recordings, behavior, perturbations and anatomy, integrative neurocybernetic models can provide statistical amplification, few-shot generalization, and mechanistic insight into shared dynamical structure, individual variation, and the control objectives that govern behavior. This agenda offers a model-centric path from fragmented data to a mechanistic science of how brains produce behavior.

  PublisherOA PDF

• Shared-AE: Automatic Identification of Shared Subspaces in High-dimensional Neural and Behavioral Activity.

  PubMed · 2025-04-01

  articleOpen accessSenior author

  Understanding the relationship between behavior and neural activity is crucial for understanding brain function. An effective method is to learn embeddings for interconnected modalities. For simple behavioral tasks, neural features can be learned based on labels. However, complex behaviors, such as social interactions, require the joint extraction of behavioral and neural characteristics. In this paper, we present an autoencoder (AE) framework, called Shared-AE, which includes a novel regularization term that automatically identifies features shared between neural activity and behavior, while simultaneously capturing the unique private features specific to each modality. We apply Shared-AE to large-scale neural activity recorded across the entire dorsal cortex of the mouse, during two very different behaviors: (i) head-fixed mice performing a self-initiated decision-making task, and (ii) freely-moving social behavior amongst two mice. Our model successfully captures both 'shared features', shared across neural and behavioral activity, and 'private features', unique to each modality, significantly enhancing our understanding of the alignment between neural activity and complex behaviors. The original code for the entire Shared-AE framework on Pytorch has been made publicly available at: https://github.com/saxenalab-neuro/Shared-AE.

  PublisherOA PDF

• Embodied sensorimotor control: computational modeling of the neural control of movement.

  PubMed · 2025-09-17

  preprintOpen accessSenior author

  We review how sensorimotor control is dictated by interacting neural populations, optimal feedback mechanisms, and the biomechanics of bodies. First, we outline the distributed anatomical loops that shuttle sensorimotor signals between cortex, subcortical regions, and spinal cord. We then summarize evidence that neural population activity occupies low-dimensional, dynamically evolving manifolds during planning and execution of movements. Next, we summarize literature explaining motor behavior through the lens of optimal control theory, which clarifies the role of internal models and feedback during motor control. Finally, recent studies on embodied sensorimotor control address gaps within each framework by aiming to elucidate neural population activity through the explicit control of musculoskeletal dynamics. We close by discussing open problems and opportunities: multi-tasking and cognitively rich behavior, multi-regional circuit models, and the level of anatomical detail needed in body and network models. Together, this review and recent advances point towards reaching an integrative account of the neural control of movement.

  PublisherOA PDF

• Multitasking Recurrent Networks Utilize Compositional Strategies for Control of Movement

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-09-16 · 1 citations

  preprintOpen accessSenior authorCorresponding

  The brain and body comprise a complex control system that can flexibly perform a diverse range of movements. Despite the high-dimensionality of the musculoskeletal system, both humans and other species are able to quickly adapt their existing repertoire of actions to novel settings. A strategy likely employed by the brain to accomplish such a feat is known as compositionality, or the ability to combine learned computational primitives to perform novel tasks. Previous works have demon-strated that recurrent neural networks (RNNs) are a useful tool to probe compositionality during diverse cognitive tasks. However, the attractor-based computations required for cognition are largely distinct from those required for the generation of movement, and it is unclear whether compositional structure extends to RNNs producing complex movements. To address this question, we train a multitasking RNN in feedback with a musculoskeletal arm model to perform ten distinct types of movements at various speeds and directions, using visual and proprioceptive feedback. The trained network expresses two complementary forms of composition: an algebraic organization that groups tasks by kinematic and rotational structure to enable the flexible creation of novel tasks, and a sequential strategy that stitches learned extension and retraction motifs to produce new compound movements. Across tasks, population activity occupied a shared, low-dimensional manifold, whereas activity across task epochs resides in orthogonal subspaces, indicating a principled separation of computations. Additionally, fixed-point and dynamical-similarity analyses reveal reuse of dynamical motifs across kinematically aligned tasks, linking geometry to mechanism. Finally, we demonstrate rapid transfer to held-out movements via simple input weight updates, as well as the generation of target trajectories from composite rule inputs, without altering recurrent dynamics, highlighting a biologically plausible route to within-manifold generalization. Our framework sheds light on how the brain might flexibly perform a diverse range of movements through the use of shared low-dimensional manifolds and compositional representations.

  PublisherOA PDFDOI

• Behavioral Classification of Sequential Neural Activity Using Time Varying Recurrent Neural Networks

  IEEE Transactions on Neural Systems and Rehabilitation Engineering · 2025-01-01

  articleOpen accessSenior author

  Shifts in data distribution across time can strongly affect early classification of time-series data. When decoding behavior from neural activity, early detection of behavior may help in devising corrective neural stimulation before the onset of behavior. Recurrent neural networks are common models for sequence data. However, standard recurrent neural networks are not able to handle data with temporal distributional shifts to guarantee robust classification across time. To enable the network to utilize all temporal features of the neural input data, and to enhance the memory of recurrent neural networks, this paper proposes a novel approach: recurrent neural networks with time-varying weights, here termed Time-varying recurrent neural networks. These models are able to not only predict the class of the time-sequence correctly, but also lead to accurate classification earlier in the sequence than standard recurrent neural networks, while also stabilizing gradient dynamics. This paper focuses on early sequential classification of spatially distributed neural activity across time using Time-varying recurrent neural networks applied to a variety of neural data from mice and humans, as subjects perform motor tasks. Time-varying recurrent neural networks detect self-initiated lever-pull behavior up to 6 seconds before behavior onset-3 seconds earlier than standard recurrent neural networks. Finally, this paper explored the contribution of different brain regions on behavior classification using SHapley Additive exPlanation value, and found that the somatosensory and premotor regions play a large role in behavioral classification.

  PublisherDOI

• Optimal feedback solutions recapitulate key features of motor cortical population dynamics

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-12-13

  articleOpen accessSenior author

  Abstract Neural populations display complex response patterns with marked transitions between distinct underlying computational strategies on very short timescales during motor tasks. Such complex-yet-structured dynamical strategies may reflect computational needs of neural systems, shaped by optimal feedback and autonomous mechanisms, in addition to biological constraints. Are there overarching computational principles that govern complex dynamical strategies exhibited by the neural population response? Here, we explore the hypothesis that computational strategies underlying neural population response represent optimal feedback solutions to the control of musculoskeletal dynamics through space for a goal. To validate this hypothesis, we develop a procedure called neural optimization using dynamical systems (NODS) learning to modify synaptic strengths within a recurrent network for locally-optimal feedback control of anatomically accurate musculoskeletal models during complex sensorimotor tasks. NODS learning works even when the objective function to be minimized is highly non-linear or the muscle model is very complex. The dynamical strategies underlying the neural network response constructed using NODS learning recapitulate key features of recorded population response. Importantly, optimal feedback solutions using NODS learning suggest that feedback mechanisms are essential for neural populations to flexibly transit between complex-yet-structured strategies. We further show that this framework provides theoretical foundations for why the solutions obtained using deep reinforcement learning algorithms extensively used to model sensorimotor tasks may explain the dynamical strategies underlying recorded population response. In summary, we develop novel methods and approaches suggesting that neural dynamics may be more strongly modulated by optimal feedback mechanisms, in addition to autonomous mechanisms, than previously appreciated.

  PublisherDOI

Frequent coauthors

• John P. Cunningham

  Columbia University

  30 shared

• Liam Paninski

  Columbia University

  22 shared

• Taiga Abe

  18 shared

• Ian Kinsella

  Columbia University

  17 shared

• E. Kelly Buchanan

  Columbia University

  14 shared

• Anne K. Churchland

  University of California, Los Angeles

  10 shared

• Simon Musall

  Forschungszentrum Jülich

  10 shared

• Mark M. Churchland

  Columbia University

  10 shared

Labs

• Saxena LabPI

Education

• PhD, Department of Electrical Engineering and Computer Sciences

  Massachusetts Institute of Technology

  2017

Awards & honors

• Sloan Research Fellowship (2025)
• Rising Stars in Electrical Engineering, UIUC (2019)
• Rising Stars in Biomedical Engineering, Johns Hopkins Univer…
• Honoree of the Graduate Women of Excellence Award, MIT (2017…