Events at Physics |
Events During the Week of August 8th through August 15th, 2021
Monday, August 9th, 2021
- No events scheduled
Tuesday, August 10th, 2021
- Network in Neutrinos, Nuclear Astrophysics, and Symmetries (N3AS) Seminar
- Probing the Dense-Matter Equation of State with Neutron Star Mergers
- Time: 2:00 pm - 3:00 pm
- Place:
- Speaker: Carolyn Raithel, Institute for Advanced Study
- Abstract: Binary neutron star mergers provide a unique laboratory for studying the dense-matter equation of state (EOS) across a wide range of parameter space, from the cold EOS during the inspiral to the finite-temperature EOS following the merger. In this talk, I will discuss the impact of the EOS on the post-merger phase of a binary neutron star coalescence, during which time the matter is heated to significant temperatures and can deviate away from its initial equilibrium composition. I will present a new set of neutron star merger simulations, which use a parametrized framework for calculating the EOS at arbitrary temperatures and compositions. I will show how systematically varying the properties of the particle effective mass affects the thermal profile of the post-merger remnant and how this, in turn, influences the post-merger evolution. Finally, I will discuss the impact of varying the slope L of the nuclear symmetry energy on the properties of the post-merger phase. In particular, I will show that the post-merger gravitational wave emission is mostly insensitive to L, but that, in contrast, the dynamical ejecta carry a weak signature of the slope of the symmetry energy.
- Host: Baha Balantekin
Wednesday, August 11th, 2021
- Physics ∩ ML Seminar
- Interpretable Deep Learning for Physics
- 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: Miles Cranmer, Princeton University
- Abstract: If we train a neural network on some dynamical system in some region of phase space, and it learns a way to execute the dynamics more efficiently than a handwritten code, how do we distill physical insight from the learned model? In this talk, I will argue that symbolic learning should play a major role in the process of interpreting a machine learning model for physical systems. I will discuss our generic method for converting a neural network that has been trained on a physical system into a symbolic model, via genetic algorithm-based symbolic regression. One of the problems with this process is working with the fact that neural networks have high-dimensional latent spaces, and genetic algorithms scale poorly with the number of features. To work around this issue, I’ll then introduce our “Disentangled Sparsity Network,” which encourages a neural network to learn an easy-to-interpret representation. I will then share several recent applications of our techniques to real physical systems, and the various insights we have discovered and rediscovered.
- Host: Gary Shiu
Thursday, August 12th, 2021
- No events scheduled
Friday, August 13th, 2021
- No events scheduled