Title:	How Language Models Work (and why that’s why they don’t)
Speaker:	Sameer Singh Associate Professor of Computer Science at University of California, Irvine

Abstract: We are at the precipice of widespread adoption of natural language processing: large language models will fundamentally change how we use and interact with devices. Through examples, we will discuss how language models can be adapted to be classifiers, summarizers, coders, writers, and conversational assistants, with little to no supervision. We will discuss the basics of neural networks, the text corpus, and the training pipeline that enables language models to behave as these general-purpose AI agents. However, we will show how this very paradigm of language modeling also introduces fundamental limitations in this technology. We will characterize these vulnerabilities in language models and discuss how they affect end-use applications. By the end of the talk, attendees will better understand the capabilities, workings, and limitations of language models.

Biography: Dr. Sameer Singh is an Associate Professor of Computer Science at the University of California, Irvine (UCI) and an AI Fellow at the Allen Institute for AI. He is working primarily on the robustness and interpretability of machine learning algorithms and models that reason with text and structure for natural language processing. Sameer was a postdoctoral researcher at the University of Washington and received his Ph.D. from the University of Massachusetts, Amherst. He received the NSF CAREER award, the UCI Distinguished Early Career Faculty award, and the Hellman Faculty Fellowship, and was selected as a DARPA Riser. His group has received funding from Allen Institute for AI, Amazon, NSF, DARPA, Adobe Research, Hasso Plattner Institute, NEC, Base 11, and FICO. Sameer has published extensively at machine learning and natural language processing venues and received conference paper awards at KDD 2016, ACL 2018, EMNLP 2019, AKBC 2020, ACL 2020, and NAACL 2022. (https://sameersingh.org/)

Title:	Medical devices cybersecurity – it’s challenges and highlights
Speaker:	Oleg Yusim Head of Product Security at Edwards Lifesciences

Abstract: Cybersecurity in medical devices requires different approach, compare to other areas. While overarchingly cybersecurity breach is synonymous with leak of sensitive data, the biggest scare for medical device is a lack of availability, closely followed by integrity of data violation. It is easy to understand why – lack of availability of the device at the moment of need, and especially in high urgency environments (such as ER) can easily lead to the death of the patient. Similarly, lack of integrity in data presented to clinician as an input into treatment course decision – or lack of integrity in the data defining medication dosage which would be administered to the patient by medical device, either of those can severely damage patient’s health all the way up to and including lethal outcome. This, and many other use case, deployment and manufacturing case specifics are what makes medical devices cybersecurity so challenging and so rewarding task to solve and the same time. We will try to touch at least on some of those topic in the upcoming talk.

Biography: Oleg comes with over 15 years of experience in Security and R&D, spanning across such areas as Avionics, Public Safety and Medical Devices. He obtained his MS in Engineering degree from University of Wisconsin-Platteville in 2017 and in 2018 relocated to Orange County with his wife and two kids to help Edwards Lifesciences to build and mature its Medical Devices Product Security program. Oleg actively participates in Health Sector Coordinating Council cross organizational taskforce groups, helping to author such papers as Joint Security Plan (rev 2) or Model Contract Language for MedTech industry, etc.