BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//UM//UM*Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/Detroit
TZURL:http://tzurl.org/zoneinfo/America/Detroit
X-LIC-LOCATION:America/Detroit
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20070311T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20071104T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260113T082021
DTSTART;TZID=America/Detroit:20260113T160000
DTEND;TZID=America/Detroit:20260113T170000
SUMMARY:Workshop / Seminar:CANCELLED - CM-AMO | Engineering topological bands in materials with chiral cavities
DESCRIPTION:Strongly coupling materials to cavity fields can affect the material’s electronic properties altering the phases of matter. In this talk\, I will first discuss the hybrid light-matter band topology of the monolayer graphene whose electrons can be coupled to both left and right circularly polarized vacuum fluctuations\, and time-reversal symmetry is broken due to an asymmetry between the two polarizations. This discussion will show how the quantum nature of photons affects the topology of the correlated light-matter hybrid wave function. A central finding of Ref. [1] is a relation between the Berry phase and the properties of exchanged photons with matter at light-matter hybridization points in the Brillouin zone. This physics turns out to be generic\, as it also emerges in stacked graphene layers [2]. I will then introduce a quantum optics model to capture the properties of the topological light-matter hybridization gaps and show the presence of chiral edge modes in these gaps [3]. For the rest of the talk\, I will describe a newly proposed photonic crystal chiral cavity [4]\, and its recent realization [5]. \n\n[1] Physical Review B 110 (12)\, L121101 (2024)\n[2] arXiv: 2504.03842\n[3] arXiv: 2510.13373\n[4] Nature Communications 16 (1)\, 5270 (2025)\n[5] arXiv: 2509.14366
UID:143295-21892653@events.umich.edu
URL:https://events.umich.edu/event/143295
CLASS:PUBLIC
STATUS:CONFIRMED
CATEGORIES:Physics,Science
LOCATION:West Hall - 340
CONTACT:
END:VEVENT
BEGIN:VEVENT
DTSTAMP:20251128T121323
DTSTART;TZID=America/Detroit:20260113T160000
DTEND;TZID=America/Detroit:20260113T170000
SUMMARY:Lecture / Discussion:Colloquium: What Kinds of Functions Do Neural Networks Learn? Low-Norm vs. Flat Solutions
DESCRIPTION:This talk investigates the fundamental differences between low-norm and flat solutions of shallow ReLU networks training problems\, particularly in high-dimensional settings. We sharply characterize the regularity of the functions learned by neural networks in these two regimes. This enables us to show that global minima with small weight norms exhibit strong generalization guarantees that are dimension-independent. In contrast\, local minima that are “flat” can generalize poorly as the input dimension increases. We attribute this gap to a phenomenon we call neural shattering\, where neurons specialize to extremely sparse input regions\, resulting in activations that are nearly disjoint across data points. This forces the network to rely on large weight magnitudes\, leading to poor generalization. Our analysis establishes an exponential separation between flat and low-norm minima. In particular\, while flatness does imply some degree of generalization\, we show that the corresponding convergence rates necessarily deteriorate exponentially with input dimension. These findings suggest that flatness alone does not fully explain the generalization performance of neural networks.
UID:142085-21889991@events.umich.edu
URL:https://events.umich.edu/event/142085
CLASS:PUBLIC
STATUS:CONFIRMED
CATEGORIES:Mathematics
LOCATION:East Hall - 1360
CONTACT:
END:VEVENT
END:VCALENDAR