FEBRUARY MEETING: Thursday, February 26, 2026 at 5:30pm (US Central)

NEW LOCATION:
ESAM Conference Room, Tech M416
2145 Sheridan Road, Evanston, IL 60208

AGENDA:
5:30pm - Meet and greet with refreshments
6:00pm - Talk with Siqiao Mu, Ph.D. Candidate, Department of Engineering Sciences and Applied Mathematics, Northwestern University

TALK TITLE:
Gradient Algorithms for Machine Unlearning

ABSTRACT:
Machine unlearning algorithms aim to efficiently remove data from a model without retraining it from scratch, in order to remove corrupted or outdated data or respect a user's "right to be forgotten." Since empirical unlearning heuristics can be unreliable, we desire "certified" machine unlearning algorithms, which are theoretically guaranteed to achieve probabilistic indistinguishability between the unlearned model and the model retrained on the retained data samples. While several works have proposed second-order unlearning algorithms, first-order methods such as gradient descent (GD) or stochastic gradient descent (SGD) algorithms are far more computationally tractable for large neural networks. We propose "Rewind-to-Delete," a first-order unlearning algorithm that is also black-box, in that it can be applied to trained models without costly precomputation. We prove certified unlearning guarantees and derive privacy-utility-complexity tradeoffs for both the GD and SGD versions.

DATA SCIENCE NIGHTS are monthly meetings featuring presentations and discussions about data-driven science and complex systems, organized by Northwestern University graduate students and scholars. Students and researchers of all levels are welcome! For more information: http://bit.ly/nico-dsn

FUTURE DATES:
Data Science Nights will be held on Thursday evenings in the winter and spring terms, with future dates on March 19, April 30, and May 28, 2026.

Related Info

Speaker:

Ágnes Horvát, Associate Professor, Department of Communication Studies, Northwestern School of Communication

Title:

The Academic Use of Social Media, LLMs and AI-Assisted Decision-Making

Abstract:

In the digital era social media and large language models (LLMs) are reshaping scholarly communication with substantial implications for visibility, publishing, and hiring. In the first part of this talk, I present our research documenting a systematic gender gap in how scientists self-promote their work on social media platforms. I then introduce follow-up survey research that investigates the mechanisms underlying this disparity and experimentally tests whether informing scholars about the gap influences their future intentions to self-promote. The second part of my talk examines the growing role of LLMs in scientific writing. Drawing on an analysis of more than 15 million biomedical abstracts, we identify abrupt vocabulary shifts consistent with LLM-assisted writing, suggesting that a substantial share of recent abstracts (at least 13.5% in 2024) has been shaped by these systems. Our findings underscore the rapid integration of LLMs into scholarly practice and raise important questions about linguistic homogenization, authorship norms, and the future of scientific communication. Finally, I present ongoing experimental research on AI-assisted decision-making. Using controlled experiments that model academic hiring as hidden-profile tasks, we compare the effects of individual AI decision aids and group-level AI facilitators on decision accuracy and participants' satisfaction with the evaluation process. Taken together, these projects illuminate how digital technologies interact with human behavior to shape whose work gains visibility, how research is written and presented, and how consequential academic decisions may be improved.

Speaker Bio:

Ágnes Horvát is an Associate Professor of Communication Studies and Computer Science (by courtesy) at Northwestern University, and director of the Lab on Innovation, Networks, and Knowledge (LINK). Her research lies in human-centered computing and network science and investigates how online spaces operate and disseminate information. Her group strives to make digital tools more efficient for scientists, entrepreneurs, and creative artists. Her recent projects investigate the use of LLMs in scientific writing and music creation, study biases in online attention to science, identify cases of collective intelligence and opportunities for improved decision-making, and develop frameworks to examine persuasion and opinion change in online discussions. Her work has been awarded an NSF CAREER, CRII, and collaborative awards as PI. Her doctoral advisees have received highly competitive prizes, including a Northwestern Presidential Fellowship and best student paper awards at international conferences. Her research has been featured recently in Nature, The New York Times, The Washington Post, Le Monde, The Economist, and Frankfurter Allgemeine Zeitung.

Location:

In person: Chambers Hall, 600 Foster Street, Lower Level
Remote option: https://northwestern.zoom.us/j/96701776160
PW: NICO26

About the Speaker Series:

Wednesdays@NICO is a vibrant weekly seminar series focusing broadly on the topics of complex systems, data science and network science. It brings together attendees ranging from graduate students to senior faculty who span all of the schools across Northwestern, from applied math to sociology to biology and every discipline in-between. Please visit: https://bit.ly/WedatNICO for information on future speakers.

Speaker:

Steven Franconeri, Professor of Psychology, Weinberg College of Arts & Sciences; Professor of Management and Organizations, Kellogg School of Management, Northwestern University

Title:

Point Taken: A gamified Intervention that Creates Enlightened Disagreements

Abstract:

Should we drop standardized testing for college or Ph.D. admissions? Allow athletes to join teams based on gender identity? When organizational and public policies bind behavior, human coexistence requires a way to determine that collective policy. Because individuals and like-minded groups have incomplete information, constrained strategies, and biased perspectives, thoughtful debate on those policies is critical. Unfortunately, those debates too often degrade into chaotic fights.

Point Taken provides a scalable solution by translating best practices in conflict resolution and critical thinking into a structured dialogue that can be learned and played in 30 minutes. In this interactive session, you'll play a short game to feel its effects.

Players replace persuasion with a common goal of discovering why they disagree. Dialogue then unfolds thoughtfully and calmly, through chains of short written reasons and responses. We've tested the game extensively in schools and organizations, and conducted a formal pilot study. All show powerful improvements in the tone and quality of debate, across longstanding and strongly-held disagreements. I’ll give background on best practices for enlightened disagreement, show how they translate to the game, ask you to play a game, and then ask for your advice on next steps.

Speaker Bio:

Steven Franconeri is leading scientist, teacher, and speaker on visual thinking, visual communication, and the psychology of data visualization. He is a Professor of Psychology in the Weinberg College of Arts & Sciences at Northwestern, Director of the Northwestern Cognitive Science Program, as well as a Kellogg Professor of Management and Organizations by Courtesy. He is the director of the Visual Thinking Laboratory, where a team of researchers explore how leveraging the visual system - the largest single system in your brain - can help people think, remember, and communicate more efficiently.

His undergraduate training was in computer science and cognitive science at Rutgers University, followed by a Ph.D. in Experimental Psychology from Harvard University, and postdoctoral research at the University of British Columbia. His work on both Cognitive Science and Data Visualization has been funded by the National Science Foundation, as well as the Department of Education, and the Department of Defense. He has received a prestigious National Science Foundation CAREER award, given to researchers who combine excellent research with outstanding teaching, and he has received a Psychonomic Society Early Career award for his research on visual thinking.

Location:

In person: Chambers Hall, 600 Foster Street, Lower Level
Remote option: https://northwestern.zoom.us/j/97198523514
PW: NICO26

About the Speaker Series:

Wednesdays@NICO is a vibrant weekly seminar series focusing broadly on the topics of complex systems, data science and network science. It brings together attendees ranging from graduate students to senior faculty who span all of the schools across Northwestern, from applied math to sociology to biology and every discipline in-between. Please visit: https://bit.ly/WedatNICO for information on future speakers.