Yale University · Wu Tsai Institute · Department of Biomedical Engineering

We study the neural control of complex, coordinated behavior. The lab builds machine-learning models of neural activity and behavior that are constrained by anatomy and physiology — using these models to predictively understand how the brain produces movement and how behavioral structure emerges from neural dynamics.

Lab website · Publications · Google Scholar · Twitter / X · Contact

• Embodied sensorimotor control. Anatomically-accurate musculoskeletal models of monkey and mouse, driven by RNNs trained with deep reinforcement learning, to ask how the motor system produces goal-directed behavior.
• Neural dynamics & state-space modeling. Switching, time-varying, and multi-regional recurrent networks for inferring latent dynamics from neural recordings.
• Multimodal behavioral representations. Autoencoders and variational methods for disentangling shared structure across neural activity, video, and behavior — across subjects and across modalities.

• Almani, Lazzari, Walker, Saxena. Embodied sensorimotor control: computational modeling of the neural control of movement. Annual Reviews in Biomedical Engineering 2026.
• Almani, Lazzari, Saxena. A goal-driven framework for elucidating the neural control of movement through musculoskeletal modeling. bioRxiv 2025. → muSim
• Yi, Dong, Higley, Churchland, Saxena. Shared-AE: Automatic Identification of Shared Subspaces in High-dimensional Neural and Behavioral Activity. ICLR 2025. → Shared-AE
• Yi, Musall, Churchland, Padilla-Coreano, Saxena. Disentangled multi-subject and social behavioral representations through a constrained subspace VAE (CS-VAE). eLife 2025. → Behavioral-feature-extraction-CS-VAE
• Zhang, Mitelut, Arpin, Vaillancourt, Murphy, Saxena. Behavioral Classification of Sequential Neural Activity Using Time-Varying RNNs. IEEE TNSRE 2025. → TV_RNN
• Zhang, Saxena. Inference of Neural Dynamics Using Switching Recurrent Neural Networks. NeurIPS 2024. → SRNN

See the full publication list →

We are part of the Center for Neurocomputation and Machine Intelligence at the Wu Tsai Institute, Yale University. We welcome inquiries from prospective PhD students, postdocs, and undergraduates interested in computational neuroscience, motor control, machine learning for neural data, or any combination of these. Reach out via the lab contact page.

_{Saxena Lab · Office 1124, Center for Neurocomputation and Machine Intelligence, Wu Tsai Institute · 100 College Street, New Haven, CT 06510}