CIS

Abstract: As the impact of artificial intelligence (AI) continues to proliferate, computer architects must assess and mitigate its environmental impact. This talk will survey strategies for reducing the carbon footprint of AI computation and datacenter infrastructure, drawing on data and experiences from industrial, hyperscale systems. First, we analyze the embodied and operational carbon implications of […]

Abstract: Transformer-based language models have undoubtedly become the dominant and favorite architecture for language generation of our time. However, although they provide impressive text quality, they tend to be hard to control. In the domain of image synthesis, on the other hand, Denoising Diffusion Models (DDM) are the dominant approach, shining with unprecedented quality and […]

Abstract: Modern machine learning methods are revolutionizing what we can do with data, from tiktok video recommendations to biomarkers discovery in cancer research. Yet, the complexity of these deep models makes it harder to understand what functions these data-dependent models are computing, and which features they detect regarding as important for a given task. In this talk, […]

Abstract: Tumors are complex and heterogeneous systems, which challenge their classification and treatment. The Silverbush lab decodes tumor heterogeneity and plasticity to understand how cancer cells transform to become more aggressive or evade treatment. We particularly examine hard-to-treat cancers with high heterogeneity, focusing on the notorious kings of heterogeneity and aggressiveness: brain cancers. To achieve this […]

Abstract: Performance of control systems depend on the nature of available measurements. Perception of edges, keypoints, landmarks and other natural semantic features make certain coarse measurements available to control systems operating in complex environments.  This talk explores two problems related to control and estimation with such coarse measurements. First, I will introduce perception contracts—an approach for analyzing visual […]

Abstract: Cooperation through safe and trustworthy communication and interaction is fundamental to how human teams accomplish complex tasks. Yet, despite significant--and sometimes revolutionary--advances in AI, we have barely begun to unlock the potential of safe, cooperative AI. This may stem from our limited understanding of how multimodal, large-scale AI models function, the one-sided nature of […]

Abstract: I will survey recent results describing the application of modern non-convex optimization methods to the problems of reconstruction attacks on private datasets (the “poison”), and the algorithmic generation of synthetic versions of private datasets that provably provide strong privacy guarantees (the “cure”). Zoom Link (if unable to attend in-person): https://upenn.zoom.us/j/97716959173

Abstract: Artificial intelligence holds enormous promise to automate our life. At the same time, AI systems can be incredibly brittle, unreliable, and biased. In this talk, I will present a few approaches towards making AI more reliable, trustworthy, and fair, drawing on recent work on uncertainty quantification and algorithmic fairness. Specifically, I will discuss how to […]