CIS

Machine learning models pretrained on internet data have acquired broad knowledge about the world but struggle to solve complex tasks that require extended reasoning and planning. Sequential decision making, on the other hand, has empowered AlphaGo’s superhuman performance, but lacks visual, language, and physical knowledge about the world. In this talk, I will present my […]

In many computer systems today, an attacker that breaks one system component can steal data from millions of users. In this talk, I will present three systems that can withstand component compromise. I will describe (1) a single sign-on system that protects user security and privacy from a compromised single sign-on server, (2) a secure-hardware-based […]

Cloud computing, which has seen significant growth over the past decade, fundamentally relies on the sharing of hardware resources among users. This approach enhances resource utilization and reduces operational costs. However, it also enables unintended information leakage through hardware side channels. Despite the threat of side-channel attacks, cloud vendors remain skeptical about the practicality of […]

Though modern machine learning has been highly successful, as we move towards more critical applications, many challenges towards building trustworthy systems, such as ensuring robustness, privacy, and fairness, arise. Ad hoc and empirical approaches have often led to unintended consequences for these objectives, thus necessitating a principled approach. Traditional solutions often require redesigning entire pipelines […]

Generative AI has led to stunning successes in recent years but is fundamentally limited by the amount of data available.  This is especially limiting in the embodied setting – where an agent must solve new tasks in new environments. In this talk, I’ll introduce the idea of compositional generative modeling, which enables generalization beyond the […]

Today, access to high-quality data has become the key bottleneck to deploying Machine Learning (ML). Often, data that is most valuable is locked away in inaccessible silos due to unfavorable incentives and ethical-legal restrictions. This is starkly evident in healthcare, where such barriers have led to highly biased and underperforming tools. In my talk, we will […]

Large Language Models (LLMs) have brought remarkable advancements to the computing industry. However, a high barrier exists between the LLMs and the vast majority of researchers and practitioners, brought by the engineering challenges with the enormous model sizes and the substantial compute requirements. In this talk, I’ll discuss my research on system innovations to democratize […]

The potential of data to transform science and society has spurred unparalleled efforts to collect it in increasingly sensitive and granular forms, which has raised a variety of societal concerns about how this data is handled and used. Though today, at-rest and in-transit encryption are standard practices, these alone are insufficient to address the security […]

ABSTRACT: Flow matching models learn a (possibly stochastic) mapping between source and target distributions. Common paradigms include diffusion models, score matching models, and continuous normalizing flows. In this talk I will first present methods for improved training of flow matching models using ideas from optimal transport. I will then show how these improved methods can […]