Have you ever wondered what truly shapes the future you’re working toward—whether in a project, passion, or everyday decision?In this interactive workshop, hosted by the Global Health Alliance in collaboration with MIT Radius, you’ll explore what drives your work and where your decisions take you. Using a simple yet powerful framework, we’ll reflect on:The futures you’re envisioning (big or small)The people your actions touchThe values behind your choicesThe moments when small decisions really matterBring a passion, project, or activity you care about, and we’ll guide you through thought-provoking questions designed to open up new perspectives and possibilities. You’ll leave with sharper insight into how your actions fit into the bigger picture and a clearer understanding of what truly matters to you.

Silent Meditation in the Chapel on Thursdays 6-8pm, open to everyone in the MIT Community. Some sessions include Guided Meditation at 6:30pm.

Time: 12:00 PMLocation: Cambridge, MA

The Neurotech 2025 symposium presents six talks by neurotechnology pioneers whose cutting-edge innovations are changing the face of neurobiological research from molecules to cognition. There will also be short talks by MIT students and postdocs. The Symposium is open to the public, but seating is limited, and registration is required. Lunch will be provided while supplies last. For more information and to register for this event, visit the Neurotech 2025 website and click "Register Here" to be sent to the Eventbrite registration page.This year’s lineup includesSPEAKERS INCLUDE:ANNA DEVOR BOSTON UNIVERSITYMICHAEL FOX HARVARD MEDICAL SCHOOL / BWHJEFF W. LICHTMAN HARVARD UNIVERSITYMADELEINE OUDIN TUFTS UNIVERSITYXIAO WANG MIT / BROAD INSTITUTEXIN YU HARVARD MEDICAL SCHOOL / MGH

Speaker: Vardan Papyan (University of Toronto)Title: Attention Sinks: A ‘Catch, Tag, Release’ Mechanism for EmbeddingsAbstract: Large language models (LLMs) often concentrate their attention on a small set of tokens—referred to as attention sinks. Common examples include the first token, a promptindependent sink, and punctuation tokens, which are prompt-dependent. Although these tokens often lack inherent semantic meaning, their presence is critical for model performance, particularly under model compression and KV-caching. Yet, the function, semantic role, and origin of attention sinks—especially those beyond the first token—remain poorly understood.In this talk, I’ll present a comprehensive investigation revealing that attention sinks catch a sequence of tokens, tag them with a shared perturbation, and release them back into the residual stream, where they are later retrieved based on the tags they carry. Probing experiments show that these tags encode semantically meaningful information, such as the truth of a statement.This mechanism persists in models with query-key normalization—where prompt-dependent, non-BOS sinks have become more common—and DeepSeek-distilled models, where it spans more heads and accounts for greater variance in the embeddings. To support future theoretical work, we introduce a minimal task that is solvable via the catch, tag, release mechanism, and in which the mechanism naturally emerges through training.Biography: Vardan Papyan is an Assistant Professor in the Department of Mathematics at the University of Toronto, cross-appointed with the Department of Computer Science. He completed his postdoctoral studies at the Department of Statistics at Stanford University, under the guidance of David Donoho, and his PhD at the Department of Computer Science at the Technion – Israel Institute of Technology, under the supervision of Michael Elad.