Agenda
Wednesday, October 6
| 9:30 AM | JAIRO DIAZ | Multipotential Colloids |
| 11:00 AM | YI SUN | Maximum likelihood for high-noise group orbit estimation and single-particle cryo-electron microscopy |
| 1:00 PM | KATE BONNEN | Seeing in Motion |
| 2:30 PM | KEITH HAWKINS | Galactic Archaeology : Uncovering the Story of Our Home with Stellar Fossils |
| 4:00 PM | SHANA CARO | The evolution of (dis)honesty and parental care(lessness) |
Abstracts and Slides
Keith Hawkins
University of Texas at Austin
Galactic Archaeology: Uncovering the Story of Our Home with Stellar Fossils
What are the processes that govern the formation, evolution and assembly of galaxies across cosmic time? This question is among the most fundamental in modern astronomy, yet the answer still eludes us to this day. The Milky Way and its local group is the optimal laboratory for answering the questions of galaxy formation because it is one of the only systems to date where we can obtain detailed and precise data on the positions, motions and chemical composition for billions of individual stars. This is the holy grail of galactic archaeology. Recent large-scale spectroscopic surveys have enabled a detailed observations of the Milky Way. In this talk, Hawkins will discuss he current work in galactic archaeology and how data from these industrial surveys in stellar astronomy have been used to dissect the structure of our galaxy. He will also explore the future of galactic archaeology through chemical cartography.
Jairo A. Díaz
Rochester Institute of Technology
Multipotential Colloids
Millions of microscale ‘building blocks’ can self-assemble into much larger structures capable of creating rich emerging properties (e.g., cells forming tissues, particles forming crystals). However, decorating the surface and controlling the shape of a simple microscale block remain challenging, requiring nanoscale precision to efficiently encode self-assembly instructions. The complexity and cost of such tasks increase when extended to millions of building blocks.
Díaz will present a novel tool where short DNA strands are used to both design shape and program binding information for colloids in bulk. The molecular precision of DNA design opens a window for rapid nanoscopic decoration, where deformations and symmetries can be programmed with ‘molecular’ resolution. Patch symmetries and DNA sequence are sorted in the liquid phase, while controlled transition to the solid phase freezes in the DNA designs for further assembly. This method enables the production of monodisperse DNA-coated faceted particles in the order of minutes.
The novel multi-DNA patchy particles hold promise to rapidly access metastable regions in phase diagrams, offering unique opportunities to extend the number of realizable self-assembled structures.
Kate Bonnen
New York University
Seeing in Motion
Humans spend much of their lives in motion, acting in the world. Thus, our visual systems must support visual perception during self-motion and visual perception for self-motion. Advances in mobile-eye-movement and body-motion tracking have enabled recordings of how we move our eyes and bodies during self-motion in natural environments. Furthermore, such recordings give us unprecedented access to the patterns of light that actually fall on our retinae during self-motion. In this talk, Bonnen will use these data to 1) describe how humans coordinate their eye movements and body motion to support stable gait patterns in difficult terrain, and 2) examine the natural statistics of retinal motion and present a set of predictions that these statistics make for the differences in retinal-motion processing across the visual field.
Yi Sun
Harvard University
Maximum likelihood for high-noise group orbit estimation and single-particle cryo-electron microscopy
View Slides
This talk will focus on the problem of recovering an unknown signal from noisy observations subject to random and unknown rotations, a model motivated by applications to single-particle cryo-electron microscopy (cryo-EM). After describing several historical approaches to this problem, Yi will present recent results on the non-convex optimization landscape of maximum likelihood estimation, the most commonly used method for this problem. In particular, these results show that the optimization properties of the landscape in the high-noise regime are determined by the algebraic structure of the rotation group.
This is talk is based on joint works with Zhou Fan, Roy Lederman, Tianhao Wang, Yihong Wu and Sheng Xu.
Shana Caro
UT Austin
The evolution of (dis)honesty and parental care(lessness)
More Events
2026 MPS Annual Meeting
ConferenceOctober 15–16, 2026
October 22–23, 2026
November 5–6, 2026
Speakers
Ofer Aharony, Aleksey Cherman, Gabriel Cuomo SISSA, Davide Gaiotto, Jaume Gomis, + 3 more