Enjoy a warm drink, connect with old friends, and meet new ones. Please bring a snack to share (store bought or homemade).Let's talk about whatever is on your mind. Hosted MS&PC volunteers. Babies & children are welcome.Simply send an email to spousesandpartners@mit.edu and let us know that you would like to join us.ONLY FOR SPOUSES, PARTNERS, AND SIGNIFICANT OTHERS.

Special Seminar with Jacob Zavatone-VethDate: Monday, February 24, 2025Time: 10:00 am – 11:00 amLocation: McGovern Seminar Room (46-3189)Title: Mechanistic identifiability in neural circuitsAbstract: One of the central goals of neuroscience is to gain a mechanistic understanding of how the dynamics of neural circuits give rise to their observed function. A popular approach towards this end is to train recurrent neural networks (RNNs) to reproduce experimental recordings of neural activity. These trained RNNs are then treated as surrogate models of biological neural circuits, whose properties can be dissected via dynamical systems analysis. How reliable are the mechanistic insights derived from this procedure? In this talk, I will discuss some of our recent efforts to disentangle when computational mechanisms are identifiable through data-driven modeling. Focusing on the simple setting of integrator circuits, I will show how mismatches can arise both due to explicit constraints imposed by architectural choices, and due to more subtle inductive biases of learning in recurrent networks. Looking to the future, I will discuss ongoing work on model evaluation procedures that focus on mechanistic recovery.Bio: Jacob Zavatone-Veth is a Junior Fellow of the Harvard Society of Fellows. His research is broadly focused on the theory of neural computation, with particular emphasis on how representations and dynamics are learned. He was first introduced to neuroscience during his undergraduate work in physics at Yale, where he studied visual motion detection and locomotor coordination in fruit flies with Damon Clark. He then came to Harvard for his Ph.D.; his doctoral work with Cengiz Pehlevan applied tools from statistical physics to investigate the structure of learned representations in natural and artificial neural networks. He is a recipient of a 2024 NIH Director's Early Independence Award, and of the 2024 American Physical Society Dissertation Award in Statistical and Nonlinear Physic

According to UN estimates, over 122 million people are displaced due to conflict, persecution, and climate change. Many more remain unaccounted for and are broadly categorized as migrants or stateless persons. These communities face complex healthcare challenges owing to changing and unstable environments and new diseases. Professor Zaman will discuss the role technology and basic science in improving their healthcare access, along with the ethical challenges these approaches present.Muhammad Zaman is the HHMI Professor of Biomedical Engineering and Director of the Center on Forced Displacement at Boston University. He is the author of We Wait for a Miracle: Health Care and the Forcibly Displaced (Johns Hopkins University Press, 2023), four other books, and over 130 peer-reviewed research articles on innovation, refugee health, and global health. He received a Guggenheim Fellowship in 2020 for his work on antibiotic resistance in refugee camps and was previously a postdoctoral fellow in cancer research at MIT.Lunch will be available at 11:45am. Please RSVP here.Contact Kate Danahy at kdanahy@mit.edu with any questions.Join our mailing list here to learn about upcoming CIS Global Research & Policy Seminars.

Join us on Monday, February 24, 2025, at 12 pm in the Nexus, Hayden Library for a talk led by Zeynep Tufekci, Turkish-American sociologist and Henry G. Bryant Professor of Sociology and Public Affairs at Princeton University.What if the Real Threat is Artificial Good-Enough Intelligence?Are we having the wrong nightmares about AI? Many worry that “artificial general intelligence” — AGI, or when machine intelligence matches or exceeds that of humans — poses a severe threat. These newly-superintelligent machines could turn on their creators, we’ve been warned, like Skynet in the apocalyptic movie Terminator.But many early predictions — fears and hopes — about how new technologies will change the world turn out to be false or misleading. Meanwhile, significant risks that arise simply from increased speed, expanded scale or lower costs that technology enables are often overlooked.A new technology does not have to outperform humans, or even be singularly sensational compared to previous technologies, to fuel extreme turbulence and instability, usually in an unforeseen manner. Cars weren’t so transformative simply because they exceeded the speed benchmark set by horses, nor were they horseless carriages.In this talk, Tufekci will examine the disruptive and even potentially catastrophic risks from Artificial Good-Enough Intelligence — AI that can do things not necessarily as well as humans but just good enough to be useful while being faster, cheaper and deployable at scale, and in ways that go beyond current concerns such as employment effects, productivity or bias.

Want to get exercise mid-day but don’t want to go outside? Join the tunnel walk for a 30-minute walk led by a volunteer through MIT’s famous tunnel system. This walk may include stairs/inclines. Wear comfortable shoes. Free.Location details: Meet in the lobby with the big mirror, right inside the Collier Memorial entrance to Stata. Location photo below.Tunnel Walk Leaders will have a white flag they will raise at the meeting spot for you to find them.Prize Drawing: Attend a walk and scan a QR code from the walk leaders to be entered into a drawing for a getfit tote bag at the end of the getfit challenge. The more walks you attend, the more entries you get. Winner will be drawn and notified at the end of April. Winner does not need to be a getfit participant.Disclaimer: Tunnel walks are led by volunteers. In the rare occasion when a volunteer isn’t able to make it, we will do our best to notify participants. In the event we are unable to notify participants and a walk leader does not show up, we encourage you to walk as much as you feel comfortable doing so. We recommend checking this calendar just before you head out! [As of Feb 12, this calendar is defaulting to the year 1899. Click "today" to be brought to the current month.]Getfit is a 12-week fitness challenge for the entire MIT community. These tunnel walks are open to the entire MIT community and you do not need to be a current getfit participant to join.

The Civil and Environmental Engineering Department at MIT will be hosting the 13th Annual Research Day event on Monday, February 24. The topic of this year’s event is Understanding the Biodiversity-Climate-Infrastructure nexus, hosted by Associate Professors Admir Masic and Serguei Saavedra.The event will consist of lightning-fast research talks and poster presentations by CEE faculty, students, and postdocs. It will also include light food and drinks during the poster presentation reception hours.Join us for an exciting event! Please RSVP here by Friday, February 14.