Events at Physics |
Events on Wednesday, February 24th, 2021
- Plasma Theory Seminar
- Special Plasma Seminar
- Plasma rotation: from mass separation applications to light manipulation
- Time: 9:00 am
- Place: Zoom Meeting
- Speaker: Renaud Gueroult, Laplace, Université de Toulouse, CNRS, Toulouse, France
- Abstract: Rotation phenomena in plasmas are found everywhere, from plasma technologies to laboratory experiments and all the way to astrophysics, and can lead to peculiar effects. In this talk, I will first discuss how the mass differential confinement properties found in a rotating plasma could enable developing mass separation technologies [1] with high upside potential for societal applications including nuclear waste cleanup and rare earth recycling. I will then show how the problem of crossed-field rotation control is intimately tied to the question of what cross-field conductivity is in a rotating plasma, and how answering this basic question is also crucial to certain alternative magnetic confinement fusion concepts [2]. Finally, as a further illustration of the peculiar effects of rotation, I will briefly touch on the basic effect of rotation on wave polarization, and how this may be of importance for pulsars' physics [3] and light manipulation applications.
[1] Gueroult R., Rax J.-M., Zweben S. J. and Fisch, N. J., Plasma Phys. Control. Fusion, 2018, 60, 014018
[2] Rax J. M., Gueroult, R. and Fisch, N. J., Phys. Plasmas, 2017, 24, 032504
[3] Gueroult R., Shi Y., Rax J.-M. and Fisch N. J., Nat. Commun., 2019, 10, 3232
Cary Forest is inviting you to a scheduled Zoom meeting
Join Zoom Meeting
Meeting ID: 977 3908 7835
One tap mobile
+13017158592,,97739087835# US (Washington DC)
+13126266799,,97739087835# US (Chicago)
Dial by your location
+1 301 715 8592 US (Washington DC)
+1 312 626 6799 US (Chicago)
+1 929 205 6099 US (New York)
+1 253 215 8782 US (Tacoma)
+1 346 248 7799 US (Houston)
+1 669 900 6833 US (San Jose)
Meeting ID: 977 3908 7835
Find your local number:
Join by SIP
97739087835@zoomcrc.com
Join by H.323
162.255.37.11 (US West)
162.255.36.11 (US East)
115.114.131.7 (India Mumbai)
115.114.115.7 (India Hyderabad)
213.19.144.110 (Amsterdam Netherlands)
213.244.140.110 (Germany)
103.122.166.55 (Australia Sydney)
103.122.167.55 (Australia Melbourne)
149.137.40.110 (Singapore)
64.211.144.160 (Brazil)
69.174.57.160 (Canada Toronto)
65.39.152.160 (Canada Vancouver)
207.226.132.110 (Japan Tokyo)
149.137.24.110 (Japan Osaka)
Meeting ID: 977 3908 7835 - Host: Cary Forest
- Physics ∩ ML Seminar
- Neural Mechanics: Symmetry and Broken Conservation Laws in Deep Learning Dynamics
- Time: 11:00 am - 12:15 pm
- Place: Online Seminar: Please sign up for our mailing list at www.physicsmeetsml.org for zoom link
- Speaker: Daniel Kunin and Hidenori Tanaka, Stanford University
- Abstract: Understanding the dynamics of neural network parameters during training is one of the key challenges in building a theoretical foundation for deep learning. A central obstacle is that the motion of a network in high-dimensional parameter space undergoes discrete finite steps along complex stochastic gradients derived from real-world datasets. We circumvent this obstacle through a unifying theoretical framework based on intrinsic symmetries embedded in a network’s architecture that are present for any dataset. We show that any such symmetry imposes stringent geometric constraints on gradients and Hessians, leading to an associated conservation law in the continuous-time limit of stochastic gradient descent (SGD), akin to Noether’s theorem in physics. We further show that finite learning rates used in practice can actually break these symmetry induced conservation laws. We apply tools from finite difference methods to derive modified gradient flow, a differential equation that better approximates the numerical trajectory taken by SGD at finite learning rates. We combine modified gradient flow with our framework of symmetries to derive exact integral expressions for the dynamics of certain parameter combinations. We empirically validate our analytic expressions for learning dynamics on VGG-16 trained on Tiny ImageNet. Overall, by exploiting symmetry, our work demonstrates that we can analytically describe the learning dynamics of various parameter combinations at finite learning rates and batch sizes for state of the art architectures trained on any dataset.
- Host: Gary Shiu
- Department Meeting
- Time: 12:15 pm - 1:15 pm
- Place: Virtual see "abstract" for connection info
- Speaker: Sridhara Dasu, Department Chair
- AIMEE N LEFKOW is inviting you to a scheduled Zoom meeting.
Topic: Department Meeting
Time: Jan 13, 2021 12:15 PM Central Time (US and Canada)
Every week on Wed, until May 12, 2021, 18 occurrence(s)
Jan 13, 2021 12:15 PM
Jan 20, 2021 12:15 PM
Jan 27, 2021 12:15 PM
Feb 3, 2021 12:15 PM
Feb 10, 2021 12:15 PM
Feb 17, 2021 12:15 PM
Feb 24, 2021 12:15 PM
Mar 3, 2021 12:15 PM
Mar 10, 2021 12:15 PM
Mar 17, 2021 12:15 PM
Mar 24, 2021 12:15 PM
Mar 31, 2021 12:15 PM
Apr 7, 2021 12:15 PM
Apr 14, 2021 12:15 PM
Apr 21, 2021 12:15 PM
Apr 28, 2021 12:15 PM
May 5, 2021 12:15 PM
May 12, 2021 12:15 PM
Please download and import the following iCalendar (.ics) files to your calendar system.
Weekly:
Join Zoom Meeting
Meeting ID: 924 9932 5588
Passcode: 337209
One tap mobile
+19292056099,,92499325588#,,,,*337209# US (New York)
+13017158592,,92499325588#,,,,*337209# US (Washington D.C)
Dial by your location
+1 929 205 6099 US (New York)
+1 301 715 8592 US (Washington D.C)
+1 312 626 6799 US (Chicago)
+1 669 900 6833 US (San Jose)
+1 253 215 8782 US (Tacoma)
+1 346 248 7799 US (Houston)
Meeting ID: 924 9932 5588
Passcode: 337209
Find your local number:
Join by SIP
92499325588@zoomcrc.com
Join by H.323
162.255.37.11 (US West)
162.255.36.11 (US East)
115.114.131.7 (India Mumbai)
115.114.115.7 (India Hyderabad)
213.19.144.110 (Amsterdam Netherlands)
213.244.140.110 (Germany)
103.122.166.55 (Australia)
149.137.40.110 (Singapore)
64.211.144.160 (Brazil)
69.174.57.160 (Canada)
207.226.132.110 (Japan)
Meeting ID: 924 9932 5588
Passcode: 337209
- Host: Sridhara Dasu