AI Innovation Institute

Abstract: Olfaction plays an essential role in everyday life. Anosmia, the loss of the sense of smell, affects approximately 20% of adults globally and its prevalence rises with age. This can lead to malnutrition, reduced ability to detect environmental hazards, and high rates of depression. The olfactory epithelium (OE) is a specialized tissue deep within the nasal cavity that primarily contains olfactory sensory neurons (OSNs), which are crucial cells that detect odorants and relay information to the olfactory neural pathway. Current clinical assessments for anosmia cannot discern whether the cause lies within the OSNs (sensorineural) or further down the pathway (central). The ability to image the OE would help assess the health of OSNs, distinguish between sensorineural and central causes of anosmia, and inform treatments. However, the OE is a small region of tissue that is non-uniform and difficult to distinguish from its surrounding respiratory epithelia (RE). This work explores the use of an FDA-approved dye, Indocyanine Green (ICG), to highlight tissue structure differences and help distinguish between OE and RE. We found that machine learning models using texture analysis information were able to classify images of the two tissue types correctly with ~80% accuracy, while a convolutional neural network was able to classify the images directly with ~90% accuracy. This advancement will enable accurate identification and mapping of the spatial distribution of OSNs within the nasal cavity, enhancing studies on olfactory degeneration and regeneration.


Bio: Skylar Suarez has a Bachelor of Science degree in Computer Science and Engineering from the University of South Florida. She spent nearly a decade as a software and hardware engineer, then decided to follow my dream of being involved in medicine and becoming a professor. She is a PhD Candidate at the Department of Biomedical Engineering at the University of Colorado Anschutz Medical Campus, where she also teaches courses in biological data analysis and machine learning. She specializes in neural engineering, particularly in the development and design of neurosurgical tools for assessing neurodegenerative disorders.


Location: New Computer Science, Room 220 OR Join by Zoom

Abstract: Recent progress in large language and vision models demonstrates how far we can go by scaling with vast internet-scale data. In contrast, physical AI, agents that perceive and act in the real world, still lags far behind. Today, both academia and industry primarily pursue generalizable physical AI by scaling up: collecting large-scale action-video datasets or training world models that enable interaction through learned environments. However, this paradigm is inherently inefficient and will soon reach a data ceiling. In this talk, I argue for a shift from scaling up to scaling out. I introduce reality world simulators, a new paradigm that converts real-world videos into diverse, interactive simulation environments. Instead of relying on more data collection, this approach expands data through structured reconstruction and recomposition, enabling both higher data efficiency and physically grounded interaction. I will present a three-pronged approach: 1) Scaling out via Digital Twins: reconstructing controllable, interactive environments from monocular videos to support diverse agent exploration. 2) Scaling out via Digital Cousins: disentangling scene structure into compositional elements to generate large-scale variations of real-world environments. 3) Scaling out via Embodied Humans: incorporating realistic human dynamics to improve safety and social compliance in robot learning. Finally, I will outline a roadmap toward building generalizable and safe physical AI systems for open-world deployment.

Bio: Dr. Wayne Wu is a postdoctoral researcher at UCLA Computer Science, working closely with Bolei Zhou, and collaborating with Trevor Darrell (UC Berkeley EECS) and Jiaqi Ma (UCLA CEE). He received his Ph.D. in Computer Science and Technology from Tsinghua University in June 2022 and was previously a visiting Ph.D. student at Nanyang Technological University. He also spent seven years in industry, where he led the research and development of products that reached more than 10 million end users worldwide. His research lies at the intersection of computer vision, robotics, and computer graphics. He focuses on developing infrastructure and methods to scale physical AI, enabling robots to work reliably and safely in the open world. He has published over 50 papers at top-tier venues including CVPR, ICCV, ICLR, NeurIPS, and ICRA, with over 9,500 citations and 10,000 GitHub stars. His work has received a CVPR Best Paper Candidate and multiple Oral, Spotlight, and Highlight presentations. He was also honored with the 2025 UCLA Chancellor's Award for Postdoctoral Research, recognizing the best postdocs at UCLA, and he was the only awardee from the School of Engineering. He serves as an Area Chair at CVPR 2026.

Location: NCS 120

University Libraries Presents

In a digital landscape flooded with hyper-realistic AI images and manipulated media, telling fact from fiction has never been harder. This workshop introduces the art of slow looking, teaching you how to pause, carefully analyze visual evidence, and critically evaluate the authenticity of what you see online.

Register here to join

Educational objectives:

1. Describe the data modalities available in All of Us (EHR, genomics, wearables) and explainhow OMOP CDM standardization enables reproducible cohort definitions across CDRversions.
2. Evaluate the role of AI-assisted coding tools in accelerating the research-to-manuscriptpipeline while maintaining investigator oversight and scientific rigor.
3. Apply a secure two-environment workflow to distinguish between exportable code artifactsand environment-bound data outputs when working within controlled-access researchplatforms.

Speaker: Hants Williams, Ph.D., R.N., Applied Health Informatics, Stony Brook University

Location: MART Building, Room: 7M-0602 (7th Floor)

Zoom: https://stonybrook.zoom.us/j/95617197636?pwd=KytzZ2pVRG9SZGpKZUtpNXJISjNjZz09
Meeting ID: 95617197636
Passcode: 924293

Join Zoom Meeting https://stonybrook.zoom.us/j/511954566?pwd=d3hGdHlGZVJISnV3d2pRcnBRbFBJZz09 (ID: 511954566, password: fxnmj4) Join by phone (US) +1 646-876-9923 Join using SIP 511954566@zoomcrc.com (passcode: 542040) Joining instructions: https://www.google.com/url?q=https://applications.zoom.us/addon/invitation/detail?meetingUuid%3DDBgsLRhCTQCGwD9DAxf7mw%253D%253D%26signature%3D02e2f3393481f8bc18cbb5520e174a6b8ceee7e9e58c9dca9d0aa4db9a63279a&sa=D&usg=AOvVaw1e3Mi7F_8TdXl5XstjsrIx Password: fxnmj4

Abstract: Many unresolved legal questions over LLMs and copyright center on memorization: whether specific training data have been encoded in the model's weights during training, and whether those memorized data can be extracted in the model's outputs. While many believe that LLMs do not memorize much of their training data, recent work shows that substantial amounts of copyrighted text can be extracted from open-weight models. However, it remains an open question if similar extraction is feasible for production LLMs, given the safety measures these systems implement. We investigate this question using a two-phase procedure: (1) an initial probe to test for extraction feasibility, which sometimes uses a Best-of-N (BoN) jailbreak, followed by (2) iterative continuation prompts to attempt to extract the book. We evaluate our procedure on four production LLMs -- Claude 3.7 Sonnet, GPT-4.1, Gemini 2.5 Pro, and Grok 3 -- and we measure extraction success with a score computed from a block-based approximation of longest common substring (nv-recall). With different per-LLM experimental configurations, we were able to extract varying amounts of text. For the Phase 1 probe, it was unnecessary to jailbreak Gemini 2.5 Pro and Grok 3 to extract text (e.g, nv-recall of 76.8% and 70.3%, respectively, for Harry Potter and the Sorcerer's Stone), while it was necessary for Claude 3.7 Sonnet and GPT-4.1. In some cases, jailbroken Claude 3.7 Sonnet outputs entire books near-verbatim (e.g., nv-recall=95.8%). GPT-4.1 requires significantly more BoN attempts (e.g., 20X), and eventually refuses to continue (e.g., nv-recall=4.0%). Taken together, our work highlights that, even with model- and system-level safeguards, extraction of (in-copyright) training data remains a risk for production LLMs.

Speaker: Xinyue

Location: CS2311

Abstract: Current Chain-of-Thought (CoT) verification methods predict reasoning correctness based on outputs (black-box) or activations (gray-box), but offer limited insight into \textit{why} a computation fails. We introduce a white-box method: \textbf{Circuit-based Reasoning Verification (CRV)}. We hypothesize that attribution graphs of correct CoT steps, viewed as \textit{execution traces} of the model's latent reasoning circuits, possess distinct structural fingerprints from those of incorrect steps. By training a classifier on structural features of these graphs, we show that these traces contain a powerful signal of reasoning errors. Our white-box approach yields novel scientific insights unattainable by other methods. (1) We demonstrate that structural signatures of error are highly predictive, establishing the viability of verifying reasoning directly via its computational graph. (2) We find these signatures to be highly domain-specific, revealing that failures in different reasoning tasks manifest as distinct computational patterns. (3) We provide evidence that these signatures are not merely correlational; by using our analysis to guide targeted interventions on individual transcoder features, we successfully correct the model's faulty reasoning. Our work shows that, by scrutinizing a model's computational process, we can move from simple error detection to a deeper, causal understanding of LLM reasoning.

Speaker: Xianjun Yang

Location: Old Computer Science Building - CS2311

Join University Libraries for an engaging panel discussion where we delve in and learn about the impacts of artificial intelligence on the 2024 US elections! Panelists are Paige Lord, Tom Costello, and Musa al-Gharbi. The discussion will be moderated by Library Dean, Karim Boughida. Co-sponsored by the Office of Diversity, Inclusion, and Intercultural Initiatives.

Please RSVP for Democracy in the Digital Age: AI's Influence on 2024 Elections here.