AI Innovation Institute

Abstract:
Quantum Machine Learning (QML) holds significant promise for solving computational challenges across diverse domains. However, its practical deployment is constrained by the limitations of noisy intermediate-scale quantum (NISQ) devices, including noise, limited scalability, and trainability issues in variational quantum circuits (VQCs). We introduce the multi-chip ensemble VQC framework, which partitions high-dimensional computations across smaller quantum chips to enhance scalability, trainability, and noise resilience. We show that this approach mitigates barren plateaus, reduces quantum error bias and variance, and maintains robust generalization through controlled entanglement. Designed to align with current and emerging quantum hardware, the framework demonstrates strong potential for enabling scalable QML on near-term devices, as validated by experiments on standard benchmark datasets (MNIST, FashionMNIST, CIFAR-10).

IACS Student Seminar Speaker:
Junghoon Park, Seoul National University
BA in Economics, Seoul National University, Korea
PhD Candidate for Interdisciplinary Programme in Artificial Intelligence at Seoul National University
Visiting Researcher at Brookhaven National Laboratory


Current Research Interests
Quantum Machine Learning


Recent Papers
Park, J., Cha, J., Chen, S. Y.-C., Yoo, S., & Tseng, H.-H. (2025). Addressing the Current Challenges of Quantum Machine Learning through Multi-Chip Ensembles. In Review at ICML.
Park, J., Kim, K., & Cha, J. (2025). How to Assess AI Ethics: Suggestions for Ethical Rating Agencies. In Review at IJCAI.
Park, J., Cha, J., Chen, S. Y.-C., Yoo, S., & Tseng, H.-H. (2024, 15-20 Sept.). Over the Quantum Rainbow: Explaining Hybrid Quantum Reinforcement Learning. 2024 IEEE International Conference on Quantum Computing and Engineering (QCE).
Park, J., Lee, E., Cho, G., Hwang, H., Kim, B.-G., Kim, G., Joo, Y. Y., & Cha, J. (2024). Gene-Environment Pathways to Cognitive Intelligence and Psychotic-Like Experiences in Children. eLife, 12, RP88117. DOI:10.7554/eLife.88117

This seminar will be held in person (food provided!) in the IACS Seminar Room, and online (zoom link below!)
https://stonybrook.zoom.us/j/96548538719?pwd=jBmI43H68q2UkdcRRjVbTkgrC6F942.1
Meeting ID: 965 4853 8719
Passcode: 493290

Abstract: Large language models (LLMs) may exhibit unintended or undesirable behaviors. Recent works have concentrated on aligning LLMs to mitigate harmful outputs. Despite these efforts, some anomalies indicate that even a well-conducted alignment process can be easily circumvented, whether intentionally or accidentally. Does alignment fine-tuning yield have robust effects on models, or are its impacts merely superficial? In this work, we make the first exploration of this phenomenon from both theoretical and empirical perspectives. Empirically, we demonstrate the elasticity of post-alignment models, i.e., the tendency to revert to the behavior distribution formed during the pre-training phase upon further fine-tuning. Leveraging compression theory, we formally deduce that fine-tuning disproportionately undermines alignment relative to pre-training, potentially by orders of magnitude. We validate the presence of elasticity through experiments on models of varying types and scales. Specifically, we find that model performance declines rapidly before reverting to the pre-training distribution, after which the rate of decline drops significantly. Furthermore, we further reveal that elasticity positively correlates with the increased model size and the expansion of pre-training data. Our findings underscore the need to address the inherent elasticity of LLMs to mitigate their resistance to alignment.

Speaker: Huajian Zhang

Location: CS2311

Abstract: Humans perceive the world around them by recognizing global patterns and structures such as object parts, branches, their spatial arrangement, and so on. Most deep learning models, however, take a fundamentally local approach. They process images pixel-by-pixel rather than focusing on structures as a whole. While these models indeed perform well on many tasks, the local (pixel-level) versus global (structure-level) disconnect makes them harder to interpret and control.

Topology, in a general sense, is a mathematical language for describing structure. It delineates how different parts of an image relate to one another, capturing both individual structures and their overall layout. Preserving topology enforces structural correctness and, by extension, semantic validity.

In this thesis, we investigate how topological constraints can be used to bridge the gap between local and global understanding. We use topology to inform the design of deep learning models that are explicitly structure-aware. Our thesis focuses on dense prediction tasks, which include image segmentation, uncertainty estimation, and generative modeling. First, we introduce a topological interaction module for semantic segmentation that encodes containment and exclusion constraints directly into the learning process. This preserves anatomical hierarchies and improves multi-class consistency. Next, since segmentation models can never be truly perfect, we address the need for reliable uncertainty estimation to identify error-prone regions. Unlike conventional pixel-wise uncertainty maps, which tend to be noisy and difficult to interpret, we propose reasoning at the level of structural units--branches and connections--which are more visually discernible and actionable. Finally, we leverage topology for generative modeling. We propose a topology-guided diffusion framework that can be controlled using structural attributes like object count and connectivity.

Together, these contributions establish a unified approach to topology-informed, structure-preserving dense prediction models. By integrating topological reasoning with deep networks, this thesis advances models that are not only accurate, but also structurally consistent, interpretable, and controllable. The results from this thesis have been published in ECCV, NeurIPS, and ICLR.

Speaker: Saumya Gupta

Location: New Computer Science (NCS) 120


Zoom: https://stonybrook.zoom.us/j/93643318604?pwd=kv8DagpbayzizivU29UCYItnlzlYRM.1&jst=2

Abstract: Drawing on group-theoretic and information-theoretic foundations, we propose information lattice learning (ILL) as a general framework to learn rules of a signal (e.g., an image or a probability distribution). In our definition, a rule is a coarsened signal used to help us gain one interpretable insight about the original signal. To make full sense of what might govern the signal's intrinsic structure, we seek multiple disentangled rules arranged in a hierarchy, called a lattice. Compared to representation/rule-learning models optimized for a specific task (e.g., classification), ILL focuses on explainability: it is designed to mimic human experiential learning and discover rules akin to those humans can distill and comprehend. We will detail the mathematical foundations and algorithms of ILL, and illustrate how it addresses the fundamental question what makes X an X by creating rule-based explanations designed to help humans understand. Our focus is on explaining X rather than (re)generating it. We show ILL's efficacy and interpretability on benchmarks and assessments, as well as a demonstration of ILL-enhanced classifiers achieving human-level digit recognition using only one or a few MNIST training examples (1-10 per class). We present applications in knowledge discovery, using ILL to distill music theory from scores and chemical laws from molecules and further revealing connections between them. We close with some early work on understanding the principles that govern scattering amplitudes in Super Yang-Mills theory, rather than just predicting them.

Biography: Lav R. Varshney is the Della Pietra Infinity Professor and inaugural director of the AI Innovation Institute at Stony Brook University. He is co-founder and CEO of Kocree, Inc., a startup company building novel human-controllable AI for discovery and creativity, and chief scientist of Ensaras, Inc., a startup company focused on AI and wastewater treatment. He holds appointments at RAND Corporation and at Brookhaven National Laboratory. He was previously on the faculty of the University of Illinois Urbana-Champaign, a visiting scholar at Northwestern's Kellogg School of Management, a principal research scientist at Salesforce Research AI, and a research staff member at IBM Research. He is a former White House staffer, having served on the National Security Council staff as a White House Fellow, where he contributed to national AI and wireless communications policy. His research interests include information theory and artificial intelligence. He received his B.S. degree from Cornell University and his S.M. and Ph.D. degrees from the Massachusetts Institute of Technology.

Location: Room 102

Spring 2025, Mondays 3.30 to 4.50 pm, NCS 220.

The seminar will be jointly taught by Prof. Chao Chen, chao.chen.1@stonybrook.edu and Prof. Dimitris Samaras samaras@cs.stonybrook.edu

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 Ph.D. program or (2) receive permission from the instructors.

Each seminar will consist of multiple short talks (around 15 minutes) by multiple students. Students can register for 1 credit for CSE656. Registered students must attend and present a minimum of 2 talks. Registered students must attend in person. Up to 3 absences will be excused. Everyone else is welcome to attend.

Join here. Meeting ID: 927 2069 8658. Passcode: 130934.
.

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.

Prof. Eugene A. Feinberg, from the Department of Applied Mathematics and Statistics, presents, Recent Developments in Markov Decision Processes Relevant to AI on April 4 at 4p. The talk discusses recent developments in Markov Decision Processes potentially relevant to artificial intelligence. These developments include complexity estimations for exact and approximate algorithms, decision making with incomplete information and multiple criteria, and continuity properties of optimal values and expectations. Dr. Eugene A. Feinberg is currently Distinguished Professor in the Department of Applied Mathematics and Statistics at Stony Brook University. He is an expert on applied probability, stochastic models of operations research, Markov decision processes, and on industrial applications of operations research and statistics. He has published more than 150 papers and edited the Handbook of Markov Decision Processes. His research has been supported by NSF, DOE, DOD, NYSTAR (New York State Office of Science, Technology, and Academic Research), NYSERDA (New York State Energy Research and Development Authority) and by industry. He is a Fellow of INFORMS (The Institute for Operations Research and Management Sciences) and has received several awards including 2012 IEEE Charles Hirsh Award for developing and implementing smart grid technologies, 2012 IBM Faculty Award, and 2000 Industrial Associates Award from Northrop Grumman. Dr. Feinberg is an Associate Editor for Mathematics of Operations Research and for Applied Mathematics Letters. He is an Area Editor for Operations Research Letters. Refreshments will be provided

Abstract: Autonomous systems, whether on Earth or in space, rely on 3D perception to understand and interact with the world around them. Yet traditional techniques for 3D understanding often depend on human designed features, fixed sensors, and conventional imaging modalities. This constrained approach can limit every stage of perception, from sensing to interpretation to decision making.
In this talk, we'll explore an alternative paradigm for imaging: physically based neural representations for 3D scenes and 3D sensing systems. We will discuss how recent advances in large scale learned representations can be used to jointly optimize both 3D scene models and the design of sensing systems for 3D capture, with the goal of enabling task specific perception systems.
Unlike modern AI models trained on internet scale datasets, these specialized 3D representations typically operate in data sparse regimes and therefore require a different kind of prior. We'll examine how grounding these learned representations in the physics of light transport can improve our understanding of scene structure, and inform imaging system design even with limited data. By connecting physical insights with learned representations, we'll highlight new possibilities for robust, efficient, and adaptive perception in challenging environments.

Speaker: Nikhil Behari is a graduate student in the Camera Culture group at the MIT Media Lab, advised by Professor Ramesh Raskar. His research interests include computational imaging, 3D scene understanding, and multi-agent decision-making under uncertainty, with a focus on automating imaging system design for 3D perception in human and planetary health. His research is supported by the NASA Space Technology Graduate Research Fellowship. He received his bachelor's in Computer Science and Statistics from Harvard University in 2022.