As we begin the 2024-25 academic year, CIS is thrilled to welcome Joel Ramirez from Stanford to our teaching faculty. He will be contributing to the CIS 2400 and 1100 courses, further enhancing our program’s academic strength. Over the past six years, Penn has made over 25 faculty hires, continuing its incredible growth. This year, we are also excited about the near completion of Amy Gutmann Hall, a new space dedicated to data science and engineering collaboration, which will host key research centers and foster innovation across multiple fields.
In addition to new spaces, Penn is launching the Penn Advanced Research Computing Center (PARCC), offering cutting-edge GPU and CPU resources for advanced research. CIS faculty are involved in groundbreaking research projects, including AI-enabled medical treatments and reducing data center energy consumption. These initiatives showcase the department’s leadership in tackling real-world challenges with innovative solutions.
This fall also marks the launch of the first Ivy League Bachelor’s degree in Artificial Intelligence, with 35 students enrolled and more courses in development. Leadership updates include Joe Devietti extending his role as Undergraduate Chair and Anindya De stepping in as Graduate Chair, both bringing fresh ideas to the department. With these exciting developments, CIS is poised for another year of growth, innovation, and academic excellence.
In celebration of AI Month, we are thrilled to extend our congratulations to Andrew Zhu, under the guidance of Chris Callison-Burch, for his remarkable achievement in receiving the prestigious 2024 NSF Graduate Research Fellowship Program (GRFP). Andrew’s dedication and innovative vision have earned him this esteemed recognition, marking a significant milestone in his academic journey.
Andrew’s research endeavors are poised to address a critical challenge in the realm of Language Model Machines or Large Language Models (LLMs), where despite their formidable capabilities, they often grapple with the issue of ‘forgetting’ when confronted with complex inquiries or tasks spanning extensive timelines. For instance, while humans effortlessly navigate through multi-step queries, LLMs encounter hurdles in retaining the original context across successive stages of information retrieval. Inspired by this gap, Andrew’s previous work, FanOutQA, shed light on the limitations of current models, showcasing the vast barrier between human accuracy and machine performance.
High-level diagram showing the settings in the FanOutQA benchmark: the LLM (robot) must answer a complex question by breaking it down (bottom). There are 3 settings we test LLMs in: closed-book (the LLM cannot use any search tools, only what it was trained on), open-book (the LLM may search Wikipedia), and evidence-provided (the LLM is given the Wikipedia pages containing the answers and must extract the correct answers).
Central to Andrew’s proposed solution is the concept of recursive delegation, a groundbreaking approach aimed at endowing LLMs with the ability to allocate intricate tasks to subordinate LLMs. This hierarchical structure allows for the decomposition of complex problems into more manageable segments, with each ‘child’ LLM equipped to handle its designated subtask. Through this cascading delegation, Andrew envisions a synergistic collaboration among LLMs, comparable to a relay race where each participant contributes to achieving the final objective.
In laying the groundwork for his research, Andrew has developed Kani, a versatile framework designed to facilitate seamless interaction between LLMs and human-written code. This innovative tool empowers researchers to experiment with different LLM architectures while interfacing with Python functions effortlessly. By democratizing access to such tools, Andrew aims to catalyze progress in the field and foster a collaborative ecosystem of exploration and discovery.
Screenshot of an early proof-of-concept recursive delegation system I developed this year. In this example, it’s helping make travel plans for a Japan trip and looking at Shinkansen (bullet train) travel times on the internet. Each node (in the top right) is an independent LLM. (green=in progress, yellow=waiting, gray=completed, blue=selected)The same as above, but with the code needed to actually use Kani. It’s designed to be easy to pick up while allowing experienced developers to customize the library extensively.An overview of what the Kani system is comprised of. It manages the chat history and provides it and user-written functions to underlying LLMs (engines) with a common interface.
Andrew’s advisor, Chris Callison-Burch, effectively summarizes the significance of his work, stating,
“Andrew Zhu’s research is at the cutting edge of programming languages, natural language processing, and artificial intelligence. His groundbreaking open-source software package, Kani, elegantly solves a major weakness of large language models (LLMs) like ChatGPT by enabling them to write code and call functions. Kani seamlessly integrates LLMs into Python programs, allowing developers to write functions that can be called by the language model. This game-changing approach combines the strengths of AI and traditional programming, enabling LLMs to interweave natural language generation with complex computations.
Andrew’s work has the potential to revolutionize how we develop AI systems, paving the way for more powerful and versatile AI applications that go beyond ChatGPT’s current abilities of understanding and generating natural language to also include the abilities to create and manipulate code.”
Chris Callison-Burch
As Andrew embarks on this ambitious journey, his work holds the promise of transforming the landscape of artificial intelligence, ushering in a new era of collaboration and problem-solving. We eagerly anticipate the insights and advancements that will emerge from his pioneering efforts, underscoring the transformative potential of AI in shaping our collective future. Once again, congratulations to Andrew Zhu on this well-deserved recognition, and here’s to a future marked by innovation, collaboration, and boundless possibilities in the realm of AI.
References:
[1] Andrew Zhu, Alyssa Hwang, Liam Dugan, and Chris Callison-Burch. 2024. FanOutQA: Multi-Hop, Multi-Document Question Answering for Large Language Models. Arxiv preprint, in submission at ACL 2024. https://arxiv.org/abs/2402.14116
[2] Andrew Zhu, Liam Dugan, Alyssa Hwang, and Chris Callison-Burch. 2023. Kani: A Lightweight and Highly Hackable Framework for Building Language Model Applications. In Proceedings of the 3rd Workshop for Natural Language Processing Open Source Software (NLP-OSS 2023), pages 65–77, Singapore. Association for Computational Linguistics. https://aclanthology.org/2023.nlposs-1.8/
Assistant Professor Gushu Li has been honored with the prestigious NSF Career Award, signaling a groundbreaking venture into the world of quantum computing. The central challenge Gushu is tackling revolves around creating programming systems that can support larger-scale quantum computers. Imagine our current systems as tailor-made for small prototypes, limiting the true potential of these advanced quantum machines.
The focus of Gushu Li’s research is on enhancing the way we write and execute programs for quantum computers. The obstacle lies in the fact that existing systems are like specialized tools for small tasks. They are not equipped to handle the complex demands of larger-scale quantum computers. To address this, Gushu is essentially upgrading the software that controls quantum computers, making it smarter and more adaptable.
The NSF Career Award doesn’t just fuel groundbreaking research but also enables Gushu Li to expand the team. New graduate students will be brought in to contribute to related projects and help build the proposed software improvements. Beyond the technical aspects, Gushu is also committed to sharing knowledge and raising awareness. The award will support the creation of two courses focused on quantum computing. They won’t just be for tech enthusiasts, but for anyone interested in understanding this cutting-edge technology.
Additionally, Gushu Li plans to be part of existing outreach programs. Introducing quantum computing concepts to students at various levels is important to expose young minds to all academic options. The interdisciplinary nature of this research, blending computer science, math, physics, and engineering, underscores its significance in advancing our understanding of how we can manipulate the very building blocks of the universe. Gushu and the team are on a path to making meaningful contributions, thanks to the support provided by the NSF Career Award.
SI Neg. 83-14878. Date: na. Grace Murray Hopper at the UNIVAC keyboard, c. 1960. Grace Brewster Murray: American mathematician and rear admiral in the U.S. Navy who was a pioneer in developing computer technology, helping to devise UNIVAC I. the first commercial electronic computer, and naval applications for COBOL (common-business-oriented language). Credit: Unknown (Smithsonian Institution)
Early Years and Academic Prowess
Born in 1906 in the bustling heart of New York City, Grace Murray Hopper emerged as a trailblazing figure in the world of computer science. Her academic journey began at Vassar College, where she graduated in 1928 with degrees in Mathematics and Physics. Hopper furthered her education with a Master’s in Mathematics from Yale in 1930, later becoming a pioneering force in the male-dominated field.
Service in the U.S. Navy during World War II
During World War II, Hopper’s commitment to service led her to join the U.S. Navy after the Pearl Harbor bombing. Trained intensively at the Midshipmen’s School for Women at Smith College, she was assigned to the Bureau of Ships Computation Project at Harvard University. Here, she worked on the IBM Automatic Sequence Controlled Calculator (MARK 1), contributing to top-secret calculations vital for the war effort, such as rocket trajectories and anti-aircraft gun tables.
Revolutionizing Computer Science: FLOW-MATIC and COBOL
Post-war, Hopper continued her groundbreaking work, transitioning to the burgeoning field of computer science. In 1953, she proposed the revolutionary idea of writing programs in words rather than symbols. This concept led to the creation of FLOW-MATIC in 1956, the first programming language using word commands and heralded as a “user-friendly language.” Her groundbreaking efforts culminated in the development of COBOL (Common Business Oriented Language) in 1959, a milestone that revolutionized programming and became the most extensively used computer language globally by the 1970s.
Legacy Beyond the Screen: Teaching and Consulting
US Navy computer ctrs. Commodre Grace Hopper in her office.(Photo by Cynthia Johnson/Getty Images)
Despite the challenges faced by women in her era, Hopper’s legacy endured. She turned down a full professorship at Vassar to continue her work with computers, eventually becoming a senior consultant at the Digital Equipment Corporation until her passing in 1992. Grace Hopper’s influence extended beyond her technological contributions, as she also dedicated herself to teaching and mentoring future generations of computer scientists at institutions such as the Moore School of Electrical Engineering at Penn and George Washington University. Her indelible mark on the digital landscape and commitment to education continue to inspire generations in the ever-evolving world of computing.
“I think we consistently…underestimate what we can do with computers if we really try.”
This year’s Halloween celebration for Penn Engineering was a great success! Held in Quain Courtyard this past Monday, October 31, students, staff, and faculty gathered for trick or treating, hot apple cider and hot chocolate, and a costume contest. There was a great turn out of students coming in between classes and during their lunch break. In addition, incredibly decorated tables from each department covered the courtyard and there was wonderful staff participation. Even some faculty members came out in costume! It was a great experience interacting with students for some spooky and relaxing fun!
Scroll down to see the department tables and awesome costumes that came out for another fun year of Halloween festivities!
On Friday, October 29, each department of the School of Engineering and Applied Science gathered together on the West Towne Lawn in the spirit of Halloween celebration!
Students, staff and faculty were able to stop by each department’s station for delicious treats and candy, Penn Engineering swag, and fun Halloween stickers and toys.
Scroll down for some amazing photos of the day, which also included a photo backdrop and Halloween tunes!
Other CIS-affiliated Halloween events include:
The Penn Society of Women Engineers Meet and Greet – Levine Lobby, October 29, 4-5pm Come take a break from studying, meet other students and enjoy some arts and crafts and insomnia cookies.
CIS Faculty and Postdoctoral Fellow Halloween TGIF – Quain Courtyard, October 29, 5-8pm There will be a Halloween Costume Contest with gift card prizes for winners! There will also be a pumpkin carving event, food, and an expanded selection of alcoholic and non-alcoholic drinks.
School of Engineering and Applied Science Dean Vijay Kumar, President Amy Gutmann, Trustee and naming donor Harlan M. Stone, and Penn Engineering Board Chair Rob Stavis at the October 1, 2021 groundbreaking for Amy Gutmann Hall to be located on the northeast corner of 34th and Chestnut Streets. Courtesy of University of Penn Almanac site
On Friday, October 1, 2021, the University of Pennsylvania’s School of Engineering and Applied Science held a groundbreaking ceremony for its new data science building and unveiled the building’s official name, Amy Gutmann Hall, honoring Penn’s President. Amy Gutmann is the eighth and longest-serving President in Penn’s history, leading the University since 2004; her term will conclude at the end of this academic year.
Amy Gutmann Hall will serve as a hub for cross-disciplinary collaborations that harness expertise, research, and data across Penn’s 12 schools and numerous academic centers. Upon completion, it will centralize resources that will advance the work of scholars across a wide variety of fields while making the tools and concepts of data analysis more accessible to the entire Penn community.
“I am thrilled Penn Engineering’s new data science building will honor Dr. Gutmann’s remarkable legacy at Penn,” said Vijay Kumar, the Nemirovsky Family Dean of Penn Engineering. “Her Penn Compact and the principles of inclusion, innovation, and impact influenced the school’s strategic priorities from which the plan for a data science building emerged. This revolutionary new facility will create a centralized home for data science research and provide collaborative and accessible space for our faculty and students, as well as the Philadelphia community.”
The 116,000-square-foot, six-floor building will be located at the northeast corner of 34th and Chestnut Streets. Planned academic features include a data science hub, the translational and outreach arm of Penn Engineering in the area of data science and artificial intelligence; research centers for new socially aware data science methodologies and novel, bio-inspired paradigms for computing; and laboratories that will develop data-driven, innovative approaches for safer and more cost-effective health care.
The impressive building is the design of executive architects Lake/Flato, with KSS Architects serving as associate architects. The building’s architecture will signify the future and the dynamic shift from the traditional to the digital. The facility is planned to be the first mass timber building in Philadelphia and will be designed sustainably.
Construction will begin in spring 2022 and is slated for completion in 2024.
