Cancelled: CHAI Research Talk: Richard Lange PhD, Assistant Professor of Computer Science, RIT

Event Image
Picture of speaker Dr. Richard Lange

Cancelled; look for the announcement rescheduling this presentation

CHAI Fall 2024 Seminar Series Research Talk

             Refreshments will be served

DATE:                  Monday, November 4, 2024, 12:00-1:00 PM

SPEAKER:          Richard Lange PhD, Assistant Professor of Computer Science, Rochester Institute of Technology

TITLE:                 Measuring neural similarity between systems that learn and adapt           

IN PERSON:       Golisano Hall (070), Room CYB-1710/1720

HOST:                 Cecilia O. Alm, PhD, Associate Director, CHAI; Professor, Department of Psychology; Joint Program Director, MS Program in Artificial Information, School of Information; Director, AWARE-AI NSF Research Traineeship Program; Director, Computational Linguistics and Speech Processing Lab

ABSTRACT:      A fundamental question for both machine learning and neuroscience is whether two neural-like systems process information in similar ways. This is the purview of Representational Similarity Analysis (RSA), a suite of methods for quantifying how (dis)similar two sets of neural representations are by presenting both systems with the same inputs and quantifying statistical relations between the representations in each system. Three challenges with RSA-style methods are that (i) they use a one-size-fits-all approach to quantifying representational similarity that is insensitive to the architecture of the systems being studied, (ii) they do not account for the behavior of the systems downstream of the measured representations, and (iii) they quantify how similar systems are now without accounting for possible learning or adaptation that could make them more similar. The first two of these challenges are met by a recent paradigm known as Neural Stitching (NS), but the third remains open. To address this, we extend NS by quantifying neural similarity before and after a controlled amount of learning is used to bring the two systems closer together. We argue that accounting for learning and adaptation is critical for any practical applications of neural similarity measures, and conclude by sketching a possible unifying framework that will bring RSA, NS, and their adaptive or learning variants together under a single roof.
BIO: Richard Lange is an Assistant Professor in the Computer Science department here at RIT, where he recently established the BONSAI Lab for Bringing Optimality into NeuroScience and AI. He is also affiliated with the Cognitive Science PhD program and the Center for Vision Science. Richard’s research seeks ways to bring AI and neuroscience together with a common mathematical language to support a common understanding of the principles of neural representations and neural information-processing. Richard is also a Rochester native, and can often be found out on a local disc golf course on warm days or on a paddle-tennis court on cold days.

NOTE:  To schedule interpreter and/or services for this event, please use https://myaccess.rit.edu.


Contact
Susan A Brightman
585-475-2509
Event Snapshot
When and Where
November 04, 2024
12:00 pm - 1:00 pm
Room/Location: CYB-1710-1720
Who

Open to the Public

CostFREE
Interpreter Requested?

No

Topics
artificial intelligence
research
student experience