BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Harvard Math - ECPv6.15.20//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Harvard Math
X-ORIGINAL-URL:https://www.math.harvard.edu
X-WR-CALDESC:Events for Harvard Math
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20250309T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20251102T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20260308T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20261101T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20270314T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20271107T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260518T080000
DTEND;TZID=America/New_York:20260522T170000
DTSTAMP:20260520T155659
CREATED:20250724T150957Z
LAST-MODIFIED:20251103T153056Z
UID:10002830-1779091200-1779469200@www.math.harvard.edu
SUMMARY:CMSA Workshop on Calabi-Yau metrics and optimal transportation
DESCRIPTION:Workshop on Calabi-Yau metrics and optimal transportation\n\nMay 18\, 2026 @ 9:00 am – May 22\, 2026 @ 5:00 pm\n\n\n\nWorkshop on Calabi-Yau metrics and optimal transportation \nDates: May 18–22\, 2026 \nLocation: Harvard CMSA \nSee the CMSA website for more details. \nOrganizers: Freid Tong\, U Toronto and Tristan Collins\, U Toronto
URL:https://www.math.harvard.edu/event/workshop-on-calabi-yau-metrics-and-optimal-transportation/
LOCATION:CMSA\, 20 Garden St\, G10\, 20 Garden Street\, Cambridge\, MA\, 02138\, United States
CATEGORIES:CMSA EVENTS
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260520T140000
DTEND;TZID=America/New_York:20260520T150000
DTSTAMP:20260520T155659
CREATED:20260429T141842Z
LAST-MODIFIED:20260429T142307Z
UID:10003163-1779285600-1779289200@www.math.harvard.edu
SUMMARY:Separation of timescales controls feature learning and overfitting in large neural networks
DESCRIPTION:To understand the inductive bias and generalization capabilities of large\, overparameterized machine learning models\, it is essential to analyze the dynamics of their training algorithms. Using dynamical mean field theory we investigate the learning dynamics of large two-layer neural networks. Our findings reveal that\, for networks with a large width\, the training process exhibits a separation of timescales phenomenon. This leads to several key observations:\n1. The emergence of a slow timescale linked to the growth in Gaussian/Rademacher complexity of the network;\n2. An inductive bias favoring low complexity when the initial model complexity is sufficiently small;\n3. A dynamical decoupling between feature learning and overfitting phases;\n4. A non-monotonic trend in test error\, characterized by a “feature unlearning” regime at later stages of training.\nJoint work with Andrea Montanari. \nZoom: https://harvard.zoom.us/j/91864143060?pwd=liDbUVYXs47QsYhxdzXYowl8vpQGy1.1
URL:https://www.math.harvard.edu/event/separation-of-timescales-controls-feature-learning-and-overfitting-in-large-neural-networks/
LOCATION:Virtually
CATEGORIES:CMSA NEW TECHNOLOGIES IN MATHEMATICS
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260520T150000
DTEND;TZID=America/New_York:20260520T160000
DTSTAMP:20260520T155659
CREATED:20260514T141636Z
LAST-MODIFIED:20260514T141636Z
UID:10003172-1779289200-1779292800@www.math.harvard.edu
SUMMARY:Quantum Simulation of Renormalized Quantum Field Theory Hamiltonians
DESCRIPTION:The most accurate ab-initio calculations of hadronic structure have been achieved through lattice QCD\, which is classically computationally expensive. Here we propose a scalable approach to computing hadronic mass spectra and constituent particle distributions using quantum computers. We outline the Renormalization Group Procedure for Effective Particles\, a renormalization scheme for quantum field theory Hamiltonians to obtain finite observables. Then we describe the Ladder Operator Block Encoding framework\, a key ingredient in efficiently calculating mass spectra of quantum field theories on quantum computers. We show explicit resource estimates for quantum simulation of renormalized Yukawa theory and QCD. \nWe look forward to seeing you there. \nIn-person only\, to be posted after the talk on https://www.youtube.com/@mathematicalpicturelanguag2715/videos
URL:https://www.math.harvard.edu/event/quantum-simulation-of-renormalized-quantum-field-theory-hamiltonians/
LOCATION:Jefferson Lab 368\, 17 Oxford St\, Cambridge\, MA\, 02138\, United States
CATEGORIES:MATHEMATICAL PICTURE LANGUAGE
END:VEVENT
END:VCALENDAR