<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style>
</head>
<body dir="ltr">
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
Hello Everyone,</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
<br>
</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
The UO Math Club will welcome a guest speaker, Sebastian Jaramillo Diaz, who will present about Research in Hamiltonian Monte Carlo, on (this)
<b>Wednesday, May 13th, 2026, Week 7, 5:00pm, University 213</b> (same room as last quarter).</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
<br>
</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
As always, there will be free pizza. </div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
<br>
</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
We hope to see all of you there.</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
<br>
</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
Here is the abstract:</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 1em 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
Hamiltonian Monte Carlo is one of the most powerful algorithms for sampling from complicated probability distributions, but most of its theory is built around Gaussian momentum. In this talk, I will explore what changes when the momentum is instead chosen from
a Laplace distribution.</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 1em 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
The main example is the exact HMC on the circle, with a von Mises target distribution. Although this is a simple geometric setting, it reveals several subtle phenomena: the dynamics are no longer governed by a smooth kinetic energy, standard contraction arguments
break down, and the behavior of nearby particles becomes highly dependent on the geometry of the target and the choice of integration time.</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 1em 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
I will discuss how shared-momentum couplings can be used to study whether HMC contracts distances in expectation, why the Laplace case requires new ideas, and what this example teaches us about the design and analysis of HMC algorithms beyond the classical
Gaussian framework. The goal is to give both intuition and mathematical structure for understanding how non-Gaussian momentum can affect mixing, stability, and efficiency.</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
Best,</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: white; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: black;">
UO Math Club.</div>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
</body>
</html>