Description:

Monday / CS Seminar
February 24th / 12:00 PM
Hybrid / Mudd 3514

Speaker
Wenqi Jiang, ETH Zurich

Talk Title
Vector-Centric Machine Learning Systems: A Cross-Stack Approach

Abstract
"Despite the recent popularity of large language models (LLMs), the transformer neural network invented eight years ago has remained largely unchanged. It prompts the question of whether machine leanring (ML) systems research is solely about improving hardware and software for tensor operations. In this talk, I will argue that the future of machine learning systems extends far beyond model acceleration. Using the increasingly popular retrieval-augmented generation (RAG) paradigm as an example, I will show that the growing complexity of ML systems demands a deeply collaborative effort spanning data management, systems, computer architecture, and ML.

I will present RAGO and Chameleon, two pioneering works in this field. RAGO is the first systematic performance study of retrieval-augmented generation. It uncovers the intricate interactions between vector data systems and models, revealing drastically different performance characteristics across various RAG workloads. To navigate this complex landscape, RAGO introduces a system optimization framework to explore optimal system configurations for arbitrary RAG algorithms. Building on these insights, I will introduce Chameleon, the first heterogeneous accelerator system for RAG. Chameleon combines LLM and retrieval accelerators within a disaggregated architecture. The heterogeneity ensures efficient serving of both LLM inference and retrievals, while the disaggregation enables independent scaling of different system components to accommodate diverse RAG workload requirements. I will conclude the talk by emphasizing the necessity of cross-stack co-design for future ML systems and the abundant of opporutnities ahead of us."

Biography
Wenqi Jiang is a final-year PhD student at ETH Zurich, advised by Gustavo Alonso and Torsten Hoefler. He aims to enable more efficient, next-generation machine learning systems. Rather than focusing on a single layer in the computing stack, Wenqi's research spans the intersections of data management, computer systems, and computer architecture. His work has driven advancements in several areas, including retrieval-augmented generation (RAG), vector search, and recommender systems. These contributions have earned him recognition as one of the ML and Systems Rising Stars, as well as the AMD HACC Outstanding Researcher Award.

Research/Interest Areas
Data management, computer systems, and computer architecture.
---
Zoom: https://northwestern.zoom.us/j/98746799161?pwd=8tJL888y1j8GrawbwOrTXKT7S9GQA4.1
Panopto: https://northwestern.hosted.panopto.com/Panopto/Pages/Viewer.aspx?id=cd3b91de-4058-478b-8d42-b28901171354
Community Connections Topic: Black Women in Computing

Description:

Wednesday / CS Seminar
February 26th / 12:00 PM
Hybrid / Mudd 3514

Speaker
Tianyu Gao, Princeton University

Talk Title
Enabling Language Models to Process Information at Scale

Abstract
Language models (LMs) are highly effective at understanding and generating text, holding immense potential as intuitive, personalized interfaces for accessing information. Expanding their ability to gather and synthesize large volumes of information will further unlock transformative applications, ranging from generative search engines to AI literature assistants. In this talk, I will present my research on advancing LMs for information processing at scale. (1) I will present my evaluation framework for LM-based information-seeking systems, emphasizing the importance of providing citations for verifying the model-generated answers. Our evaluation highlights shortcomings in LMs’ abilities to reliably process long-form texts (e.g., dozens of webpages), which I address by developing state-of-the-art long-context LMs that outperform leading industry efforts while using a small fraction of the computational budget. (2) I will then introduce my foundational work on using contrastive learning to produce performant text embeddings, which form the cornerstone of effective and scalable search. (3) In addition to building systems that can process large-scale information, I will discuss my contributions to creating efficient pre-training and adaptation methods for LMs, which enable scalable deployment of LM-powered applications across diverse settings. Finally, I will share my vision for the next generation of autonomous information processing systems and outline the foundational challenges that must be addressed to realize this vision.

Biography
Tianyu Gao is a fifth-year PhD student in the Department of Computer Science at Princeton University, advised by Danqi Chen. His research focuses on developing principled methods for training and adapting language models, many of which have been widely adopted across academia and industry. Driven by transformative applications, such as using language models as information-seeking tools, his work also advances robust evaluation and fosters a deeper understanding to guide the future development of language models. He led the first workshop on long-context foundation models at ICML 2024. He won an outstanding paper award at ACL 2022 and received an IBM PhD Fellowship in 2023. Before Princeton, he received his BEng from Tsinghua University in 2020. 

Research/Interest Areas
Natural language processing, language models
---
Zoom: https://northwestern.zoom.us/j/91611540710?pwd=7yBeDMdu6jAcoK2wFQj9Pal31bxk6K.1
Panopto: https://northwestern.hosted.panopto.com/Panopto/Pages/Viewer.aspx?id=c651c906-f33f-48e5-87f1-b2890117aac0
Community Connections Topic: Supporting First Generation Students

Description:

Wednesday / CS Seminar
March 5th / 12:00 PM
Hybrid / Mudd 3514

Speaker
Jas Brooks, University of Chicago

Talk Title
Expanding Human & Computer Senses through Perceptual Engineering

Abstract
"Imagine a future where sensory experiences are as easily customizable as adjusting phone settings—reducing sweetness to encourage healthier eating, modulating perceived temperature for comfort, or extending sensory range to detect imperceptible noxious gases. Despite the transformative potential of such advancements, today’s computer interfaces struggle to integrate rich and intimate senses like temperature, touch, taste, and smell due to persistent challenges such as power inefficiency, miniaturization difficulties, and the inability to target specific sensory effects.
I argue that entirely new interfacing techniques are needed. To address these barriers, I focus on perceptual engineering—the design and implementation of interfaces that precisely alter sensory mechanisms to systematically change perception in a controlled and reproducible manner. My research first explores this through chemical interfaces, a new class of wearable systems that induce sensory feedback by interacting directly with the body’s chemical pathways. Unlike traditional mechanical stimulation or sensory substitution, chemical interfaces are power-efficient, versatile, and selective: they reduce energy consumption for temperature feedback (CHI’20 Best Paper), create diverse haptic sensations with a single miniaturized actuator (UIST’21), and precisely modify taste, such as reducing sweetness perception to promote healthier diets (UIST’23 Demo Honorable Mention).
However, perceptual engineering extends beyond chemical interactions. My work demonstrates that this approach generalizes across multiple stimulation modalities—from electrical stimulation of the septum to evoke smell-like sensations (CHI’21) to thermal modulation of the nose that alters perceived airflow (UIST’24). These interfaces not only overcome technical limitations but also open new possibilities in health, training, and immersive experiences. For example, taste retargeting offers a novel approach to improve eating habits (UIST’23), stereo-smell could enable users to detect and localize harmful gases in high-risk environments (CHI’21), and interfaces that make one feel like they are breathing more air than they actually inhale could support health interventions like anxiety management or improved face mask compliance (UIST’24).
Perceptual engineering lays the foundation for the future I envision where users can actively shape their perceptions to improve health, enhance comfort, and enrich their interactions with both digital and physical environments."

Biography
Jas Brooks (they/them) is a Computer Science Ph.D. candidate at the University of Chicago. Their research reimagines how technology integrates with human senses—temperature, touch, taste, and smell—by focusing on perceptual engineering, a framework for designing technologies that precisely fine-tune sensory perception by combining methods from computer science, neurobiology, and psychophysiology. Jas’s research has been published at top-tier HCI venues such as ACM CHI and UIST, earning two Best Paper Awards, and has been recognized with honors like the 2023 Rising Star in EECS, 2024 Siebel Scholar distinction, and an NSF Graduate Research Fellowship. Their work has attracted media coverage from outlets like WIRED and Fast Company. Beyond their doctoral work, Jas studies and conserves historical scent technologies like AromaRama and Smell-O-Vision, documents early 20th-century scent-enhanced media, and curates exhibitions bridging historical and modern olfactory practices.

Research/Interest Areas
Human-Computer Interaction; Embedded Systems; Wearable Computing; Bio-Inspired Computing
---
Zoom: https://northwestern.zoom.us/j/95683661668?pwd=HyBG6Stkwd7E68xHqrwNFh17mt5Tsr.1
Panopto: https://northwestern.hosted.panopto.com/Panopto/Pages/Viewer.aspx?id=2a308f86-ce8c-44cc-8566-b2890118aeaa
Community Connections Topic: Intersectionality and Identity