Patricio Salvatore La Rosa, Ph.D. is a data science and AI leader at Bayer Crop Science with prior analytics leadership experience at Monsanto. He also serves as Adjunct Faculty and an Industrial Advisory Board member at Washington University in St. Louis, where he earned his M.S. and Ph.D. in Electrical Engineering and completed postdoctoral training in Biostatistics at the WashU School of Medicine.

Tanner R. Cooper is an Associate and Senior Building Performance Consultant at IMEG, specializing in building analytics, energy optimization, and commissioning services. He holds both a B.S. and M.S. in Mechanical Engineering from Washington University in St. Louis and applies data-driven strategies to improve sustainability and operational efficiency in the built environment.

View on site | Email this event

Recent advances in neuroscience demand hardware platforms capable of recording and modulating neural activity at increasingly large scales. In this seminar, I will present our work on integrated circuits for high-density neural recording and minimally invasive stimulation. The talk will highlight a 5,376-channel neural recording ASIC enabling simultaneous full-band acquisition with compact stacked packaging and high-speed data streaming. I will also present two magnetic-field-based stimulation approaches: one leveraging ferromagnetic resonance of magnetic nanoparticles for efficient localized heating, and another using a miniaturized magnetic coil for energy-efficient neural excitation. These systems demonstrate scalable and minimally invasive strategies for neuromodulation. Finally, I will briefly discuss the broader vision of integrated bioelectronic platforms that extend neural interfaces beyond conventional electrical modalities.

View on site | Email this event

Sequential decision making by a large set of myopic agents has gained significant attention over the past decade. In such settings, even a little amount of experimentation from a few agents would benefit all others but obtaining such experimentation could be challenging for a central planner. The academic literature has focused on mechanisms for promoting experimentation through monetary incentives and persuasion through careful information disclosure. We study simple controls that the central planner can use to coordinate experimentation. We consider a set of myopic agents that observe their own histories but not the histories of other agents. In a continuous-time stochastic multi-armed bandit model, the agents pick arms myopically and receive instantaneous rewards. Meanwhile, the central planner can observe the history of all agents. We consider a class of policies where the central planner is allowed to irrevocably remove arms. We show that an appropriately chosen policy within this class can generate the needed experimentation and match the regret bounds for a centralized problem thus mitigating the cost of decentralization. We also quantify the minimum number of agents that are needed for such a policy to be asymptotically optimal and the impact of the number of agents on the speed of learning. We then extend our study to online stochastic linear programs and characterize a geometric policy that mitigates the cost of decentralization and myopia.

View on site | Email this event

As AI/ML pervades the scientific community, scientists and engineers need to trust that these models and methods are human aligned while appropriately obeying fundamental laws of nature. In this talk, trustworthiness for scientific AI/ML is presented along three dimensions: philosophical, computational, and experimental. Science-oriented trustworthy AI frameworks provide a philosophical approach to human alignment in the laboratory; these frameworks evaluate AI/ML for properties such as interpretability and fairness in ways that are meaningful to the research community, and I discuss recent progress towards community-based trustworthiness standards. I then present a case study where computational approaches leverage novel explainable uncertainty quantification (xUQ) methods to evaluate an AI/ML model that fails to generate trustworthy predictions. Finally, because physical models benefit from physical validation, I present progress towards implementing proposed xUQ methods in an autonomous laboratory experiment for battery testing. These perspectives provide an outlook for how trustworthy AI/ML can support future research endeavors.

View on site | Email this event

Imagine trying to reverse-engineer how a car works by watching only the spark plugs fire, ignoring everything else under the hood. This talk is about the rest of the engine. I will present recent theoretical and computational work aimed at explaining growing experimental evidence that non-neuronal cells (specifically astrocytes) play active roles in behavior, learning, and memory.

Across circuit- and systems-level models, we show that neuron–astrocyte interactions can support powerful associative memory mechanisms (e.g., modern Hopfield networks and transformers) and generate stable, autonomous behavior in virtual robotic models of organisms (larval zebrafish) without external rewards. These latter models reproduce key behavioral transitions and quantitatively predict large-scale neural and astrocytic data observed in freely behaving animals. The central message of this talk is simple: reverse-engineering natural intelligence will likely require understanding not just neurons, but the rest of the engine.

View on site | Email this event