AI Innovation Institute

Abstract: Jailbreak attacks circumvent LLMs' built-in safeguards by concealing harmful queries within adversarial prompts. While most existing defenses attempt to mitigate the effects of adversarial prompts, they often prove inadequate as adversarial prompts can take arbitrary, adaptive forms. This paper introduces RobustKV, a novel jailbreak defense that takes a fundamentally different approach by selectively removing critical tokens of harmful queries from key-value (KV) caches. Intuitively, for an adversarial prompt to be effective, its tokens must achieve sufficient `importance' (measured by attention scores), which consequently lowers the importance of tokens in the concealed harmful query. Therefore, by carefully evicting the KVs of low-ranked tokens, RobustKV minimizes the harmful query's presence in the KV cache, thus preventing the LLM from generating informative responses. Extensive evaluation using benchmark datasets and models demonstrates that RobustKV effectively counters state-of-the-art jailbreak attacks while maintaining the LLM's performance on benign queries. Notably, RobustKV creates an interesting effectiveness-evasiveness dilemma for the adversary, leading to its robustness against adaptive attacks.

Speaker: Tanqiu Jiang

Where: NCS 220 and Zoom (https://stonybrook.zoom.us/j/6406956411)

Discover how U.S. Census Bureau Tools can help you find free data for your research projects, community, and more. See how to access the latest American Community Survey and 2020 Census data for various geographies including New York City and Long Island at data.census.gov. Learn about Community Resilience Estimates and how to navigate My Community Explorer; an interactive map-based tool which highlights demographic and socioeconomic data that measure inequality. This session will involve live demonstrations and hands-on exercises for participants. Registrants will receive the Zoom link one day prior to the event.

Please Register for SBU Libraries' AI Club: Exploring Census Data here.

Abstract:

What is the nature of linguistic knowledge, and how is it acquired from limited data? In recent years, the program of subregular linguistics has identified formal language classes expressive enough to account for most phenomena in natural language but also sufficiently limited to be efficiently learned from positive data. An advantage to these formal learning algorithms is that they come with mathematically proven guarantees about their performance, and it is easy to reason about how and why they behave the way they do.

In this talk, I discuss the Multi Tier-based 2-Strictly Local Inference Algorithm (MT2SLIA), which probably learns the syntactically relevant class of 2-Factor Muti Tier-based Strictly Local (2FMSTL) tree languages. This algorithm efficiently learns from a polynomially-sized sample of positive data by identifying missing substructures and generalizing these as constraints over tiers in a principled manner.

I will introduce a working prototype implementation of this algorithm and demonstrate its behavior on a curated sample of natural language data to show how it can learn relevant syntactic patterns.

Bio:

Logan Swanson is a third year PhD student in the Department of Linguistics at Stony Brook University. He is advised by Dr. Jefferey Heinz and Dr. Thomas Graf. His interests include learning theory, computational syntax, and language change. His current research focuses on understanding the learning-theoretic elements of natural language by designing, implementing, and testing learning algorithms for linguistically relevant formal language classes.

*Please note: this seminar will be held in person (IACS Seminar Room w/ food provided) and online.

Join Zoom Meeting
https://stonybrook.zoom.us/j/95707958315?pwd=6ITUJ0ffCXjRJb4wpt0KMDTApfSLZ0.1

Meeting ID: 957 0795 8315
Passcode: 920473

The overall purpose of this seminar is to bring together people with interests in Computer Vision theory and techniques and to examine current research issues. This course will be appropriate for people who already took a Computer Vision graduate course or already had research experience in Computer Vision. To enroll in this course, you must either: (1) be in the PhD program or (2) receive permission from the instructors.

Each seminar will consist of multiple short talks (around 10 minutes) by multiple people. Students can register for 1 credit for CSE 656. Registered students must attend and present a minimum of 2 or 3 talks. Everyone else is welcome to attend. Fill in https://forms.gle/pCVXovgfMfQwGqG38 to subscribe to our mailing list for further announcement.

This workshop focuses on how to use AI Deep Research for investigating a topic. Go from a basic search to utilize Gemini and Perplexity to find information. We will show you our steps to evaluate and gain deeper insights from the results.

Register here for the online session.

University Libraries Present: Qualitative data can be challenging to analyze and interpret effectively. In this workshop, SBU Libraries' Data Literacies Lead, Ahmad Pratama will show you how to extract meaningful insights from textual data, including understanding sentiment trends. Learn to explore qualitative data with Python using word clouds, basic natural language processing (NLP) techniques, and lexicon-based sentiment analysis with VADER.
https://stonybrook.zoom.us/meeting/register/k0r6mPYCRayk2AOGmyd0qw#/registration

CSE 600 Seminar Series | Fall 2025


Abstract: Large reasoning models have demonstrated capabilities to solve competition-level math problems, answer deep research questions, and address complex coding needs. Much of this progress has been enabled by scaling of data: pre-training data to learn vast knowledge, fine-tuning data to learn natural language reasoning, and RL environments to refine that reasoning. In this talk, I will describe the current LLM reasoning paradigm, its boundaries, and the future of LLM reasoning beyond scaling. First, I will describe the state of reasoning models and where I think scaling can lead to some additional (though perhaps limited) successes. I will then shift to discussing more fundamental issues with models that scale will not resolve in the next few years. I will touch on four current limitations: outdated knowledge, generator-validator gaps, limited creativity, and poor compositional generalization. In all cases, fundamental limitations of LLMs or of supervised learning in general make these problems challenging, inviting future study and novel solutions beyond scaling.

Bio: Greg Durrett is an associate professor in the Department of Computer Science and the Center for Data Science at New York University. His research is broadly in the areas of natural language processing and machine learning. Currently, his group's focus is on reasoning about knowledge in text, verifying correctness of generation methods, and studying how to make progress on problems that defy LLM scaling. He is a 2023 Sloan Research Fellow and a recipient of a 2022 NSF CAREER award. He has served in numerous roles for ACL conferences, recently as a member of the NAACL Board since 2024 and as Senior Area Chair for ACL 2025 and EMNLP 2025. He received his BS in Computer Science and Mathematics from MIT and his PhD in Computer Science from UC Berkeley, where he was advised by Dan Klein.

Abstract:
Coarse grained (CG) models alleviate the drawbacks of all-atom simulations. The latter still pose challenges because they are computationally expensive and give access to limited spatiotemporal scales, despite the use of modern high-performance computing clusters. CG models ignore some of the atomistic degrees of freedom, leading to fewer interatomic interactions, hence less computing time. Introducing such models emphasizes the need to properly manage these multiple scales, by carefully deriving potentials and reconstructing conformations from their CG representations, usually with the help of Machine Learning. Following a bottom-up and force matching approach, we train a Physics-Informed Neural Network to extract the CG force field parameters from all-atom simulation data. We verify our approach by applying it to fibrin monomers to study multiple-fibrin polymerization in solution at the microsecond scale, after modifying the force field to incorporate further non-bonded interactions, not present in the training data. Access to these scales will allow us to study the effects of some of the molecules' components. Furthermore, we modify recent solutions in data-driven protein backmapping. Taking advantage of the developments in graph neural networks and variational inference, we introduce an intermediate step in the all-atom reconstruction of a molecule given its CG configuration, in an attempt to more accurately de-coarsen structures whose atom-to-CG-beads ratio is very high. The combined effect of our new forward and inverse coarse graining methodology will enable the in silico study of many phenomena that are highly dynamic and intrinsically multiscale.

Bio:
Georgios Kementzidis is a third year PhD student in the Department of Applied Mathematics and Statistics at Stony Brook University. His advisor is Dr. Yuefan Deng. His research interests lie at the intersection of Computational Science, molecular dynamics (MD) simulations, and Machine Learning (ML) applications to Computational Biophysics. He is particularly interested in coarse-graining and multi-scale simulations.

*Note: this seminar will be held in-person (food provided on a first-come, first serve basis) and online*

Join Zoom Meeting https://stonybrook.zoom.us/j/99510099036?pwd=EyowuLBGvUVLZDBlG6F6chkMICFOZ7.1
Meeting ID: 995 1009 9036
Passcode: 132419

CG Group member (and SBU faculty) Chao Chen will speak on Fri, March 12, about the use of topological data analysis in machine learning for image analysis.
Chao has shared some of his research with the CG Group previously, and this will be a great opportunity to learn more about this exciting research area related to computational geometry/topology!

Time: Friday, March 12, 2pm-3pm
Place: Zoom
https://stonybrook.zoom.us/my/profweizhu?pwd=RjVIVXg3YUhudzZZQ3pheHUydTJBUT09



Title: Learning with Topological Information - Image Analysis and Label Noise
Speaker: Prof. Chao Chen (SBU)

Abstract: Modern machine learning faces new challenges. We are
analyzing highly complex data with unknown noise. Topology provides
novel structural information to model such data and noise. In this
talk, we discuss two directions in which we are using topological
information in the learning context. In image analysis, we propose a
topological loss to segment and to generate images with not only
per-pixel accuracy, but also topological accuracy. This is necessary
in analysis of images of fine-scale biomedical structures such as
neurons, vessels, etc.  Extracting these structures with correct
topology is essential for the success of downstream
analysis. Meanwhile, we discuss how to use topological information to
train classifiers robust to label noise. This is important in practice
especially when we are using deep neural networks which tend to
overfit noise. These results have been published in NeurIPS, ECCV,
ICML and ICLR.