Events

Hide Filters Show Filters
Hide Filters Show Filters

Event Type

Impact Area

Research Area

Seminar Series

Reset Filters

October 13, 2024

No events scheduled

October 16, 2024

October 17, 2024

Institutions affect preferences: The value of autonomy under liberal and authoritarian regimes

Katrin Schmelz
Santa Fe Institute
1:00P
- 2:00P

Location

Seminar Room G882 (Hewlett Room)
Add to Calendar 2024-10-17 13:00:00 2024-10-17 14:00:00 America/New_York Institutions affect preferences: The value of autonomy under liberal and authoritarian regimes ZOOM: https://mit.zoom.us/j/91586451836ABSTRACT:(joint work with Samuel Bowles and Anthony Ziegelmeyer)Do liberal societies cultivate and sustain the social norms and other preferences required to perpetuate their basic institutions in the long run? To explore the dynamic stability of liberalism as a culture-institutions coevolutionary process, we measure the value of personal autonomy and aversion to being controlled by others, among East Germans raised under Communist rule and East Germans raised in liberal Germany. Consistent with a positive response to our motivating question, in behavioral experiments we find that those raised in greater freedom adopted more pro-personal-autonomy preferences, a result unlikely to reflect age effects (based on evidence from West German cohorts). This pattern is also evident in an entirely different domain, i.e., voluntary as opposed to mandatory covid policies, and also has implications for the design of effective climate policies.BIO:Katrin is a behavioral economist and psychologist who studies how incentives and controls affect individual motivation and the influence of culture and institutions on behavior. Her recent interest is in how these mechanisms can be applied to public health and climate policies. At SFI, she draws on the complexity sciences to better understand how the interaction between institutions and people’s preferences can be taken into account when designing public policies, including measures to address climate change.She holds a postgraduate degree in psychology. Her PhD dissertation in economics at the Max Planck Institute for Economics in Jena received the Heinz Sauermann Prize of the GfeW (German Association for Experimental Economic Research), she received the State Prize of Courageous Science for her research and public engagement with respect to covid policies, and the Science Prize of the Messmer Foundation for her current behavioral research on climate policies.Three of her recent papers on how anti-COVID-19 policies change citizens’ preferences and beliefs appeared in PNAS. She has also published in Experimental Economics, The Journal of Neuroscience, and Human Brain Mapping. Her commentaries on public policy have appeared in the Washington Post, Nature News Feature, Science Insider, The Guardian, Times of India, Newsweek Japan, Radio France, VoxEU, LSE COVID-19 blog, and many other media outlets around the world.Following her Omidyar and EPE Postdoctoral Fellowship at SFI, Katrin will become a professor at the Section for Human Behavior at the Technical University of Denmark (DTU). Seminar Room G882 (Hewlett Room)

Bounded Arithmetic Reading Group

3:00P
- 5:30P

Location

Seminar Room G575
Add to Calendar 2024-10-17 15:00:00 2024-10-17 17:30:00 America/New_York Bounded Arithmetic Reading Group Bounded arithmetic is a collective name of fragments of Peano arithmetic that are closely related to complexity classes, such as P, PH, and L. The study of bounded theories is relevant to the program of showing unprovability of complexity conjectures; moreover, the techniques in this field are found useful in complexity theory and cryptography. In this reading group, we will learn basics of bounded arithmetic from scratch (without assuming backgrounds in logic).Organizer: Jiatu Li Seminar Room G575

Aligning Robot and Human Representations

Andreea Bobu
MIT Aero Astro / CSAIL
4:00P
- 5:00P

Location

32-G449 (Stata Center, Patil/Kiva Conference Room)
Add to Calendar 2024-10-17 16:00:00 2024-10-17 17:00:00 America/New_York Aligning Robot and Human Representations Abstract: To perform tasks that humans want in the world, robots rely on a representation of salient task features; for example, to hand me a cup of coffee, the robot considers features like efficiency and cup orientation in its behavior. Prior methods try to learn both a representation and a downstream task jointly from data sets of human behavior, but this unfortunately picks up on spurious correlations and results in behaviors that do not generalize. In my view, what’s holding us back from successful human-robot interaction is that human and robot representations are often misaligned: for example, our assistive robot moved a cup inches away from my face -- which is technically collision-free behavior -- because it lacked an understanding of personal space. Instead of treating people as static data sources, my key insight is that robots must engage with humans in an interactive process for finding a shared representation for more efficient, transparent, and seamless downstream learning. In this talk, I focus on a divide and conquer approach: explicitly focus human input on teaching robots good representations before using them for learning downstream tasks. This means that instead of relying on inputs designed to teach the representation implicitly, we have the opportunity to design human input that is explicitly targeted at teaching the representation and can do so efficiently. I introduce a new type of representation-specific input that lets the human teach new features, I enable robots to reason about the uncertainty in their current representation and automatically detect misalignment, and I propose a novel human behavior model to learn robust behaviors on top of human-aligned representations. By explicitly tackling representation alignment, I believe we can ultimately achieve seamless interaction with humans where each agent truly grasps why the other behaves the way they do.Bio: Andreea Bobu is an Assistant Professor at MIT in AeroAstro and CSAIL. She leads the Collaborative Learning and Autonomy Research Lab (CLEAR Lab), where they develop autonomous agents that learn to do tasks for, with, and around people. Her goal is to ensure that these agents' behavior is consistent with human expectations, whether they interact with expert designers or novice users. She obtained her Ph.D. in Electrical Engineering and Computer Science at UC Berkeley with Anca Dragan in 2023. Prior to her Ph.D. she earned her Bachelor’s degree in Computer Science and Engineering from MIT in 2017. She was the recipient of the Apple AI/ML Ph.D. fellowship, is a Rising Star in EECS and an R:SS and HRI Pioneer, and has won best paper award at HRI 2020 and the Emerging Research Award at the International Symposium on the Mathematics of Neuroscience 2023. Before MIT, she was also a Research Scientist at the AI Institute and an intern at NVIDIA in the Robotics Lab. 32-G449 (Stata Center, Patil/Kiva Conference Room)

October 21, 2024

Luca Carloni: Collaborative System-on-Chip Design with Open-Source Hardware Platforms

Luca Carloni
Columbia University
3:00P
- 4:00P

Location

G882
Add to Calendar 2024-10-21 15:00:00 2024-10-21 16:00:00 America/New_York Luca Carloni: Collaborative System-on-Chip Design with Open-Source Hardware Platforms Abstract:Open-source hardware holds a promise of sustaining the progress of the semiconductor industry in the age of heterogeneous computing. It can enable design reuse, foster design collaboration, and support workforce development. ESP is an open-source research platform for system-on-chip (SoC) design that combines a modular architecture and an agile design methodology. The ESP architecture simplifies the design and prototyping of heterogeneous chips with multiple RISC-V processor cores and dozens of loosely-coupled accelerators, all interconnected with a scalable network-on-chip. The ESP methodology promotes system-level design while accommodating different specification languages and design flows. ESP's capabilities have allowed a small team of mostly graduate students to realize two SoCs of growing complexity, each in the span of just a few months. Conceived as a heterogeneous system integration platform and developed through years of teaching at Columbia University, ESP is intrinsically suited to advance collaborative engineering across the open-source hardware community.Bio:Luca Carloni is professor and chair of Computer Science at Columbia University in the City of New York. He holds a Laurea Degree Summa cum Laude in Electronics Engineering from the University of Bologna, Italy, and the MS and PhD degrees in Electrical Engineering and Computer Sciences from the University of California, Berkeley. His research interests include heterogeneous computing, system-on-chip platforms, embedded systems, and open-source hardware. He has co-authored over one hundred and ninety refereed papers. Luca received the NSF CAREER Award, the Alfred P. Sloan Research Fellowship, and the ONR Young Investigator Award. He is an IEEE Fellow.

October 24, 2024

Thesis Defense, H.J. Terry Suh. Leveraging Structure for Efficient and Dexterous Contact-Rich Manipulation

H.J. Terry Suh
MIT CSAIL, Robot Locomotion Group
10:00A
- 11:30A

Location

32-D463 (Star)
32-D463 (Star)
Add to Calendar 2024-10-24 10:00:00 2024-10-24 11:30:00 America/New_York Thesis Defense, H.J. Terry Suh. Leveraging Structure for Efficient and Dexterous Contact-Rich Manipulation Thesis Defense: Leveraging Structure for Efficient and Dexterous Contact-Rich ManipulationSpeaker: H.J. Terry Suh In-person Location: 32-D463 (Star)Zoom Link: https://mit.zoom.us/j/7652207066Abstract: Contact-Rich Manipulation has proved challenging due to the need to consider multiple combinatoric possibilities of making or breaking contact with the surrounding environment. As a result, existing methods have often resorted to combinatorial optimization that utilizes dynamics structure but considers all possibilities exhaustively, or compute-heavy and inefficient sampling methods that utilize blackbox optimization such as Reinforcement Learning (RL). In this thesis, I aim to show that by combining structured contact smoothing in conjunction with local gradient-based control and sampling-based motion planning, we can bypass the combinatorial explosion of contact modes while still leveraging structure and achieving highly efficient contact-rich manipulation. Foremost, I shed light on how RL abstracts contact modes and optimizes difficult landscapes by combining stochastic smoothing and zeroth-order optimization; yet, I show how following a similar stochastic strategy while using gradients suffers from several drawbacks such as empirical bias and high variance. To leverage structure in a more helpful manner, I propose a method for smoothing contact dynamics without relying on stochastic smoothing, bypassing these drawbacks. Using this smoothing scheme, I show that we can achieve highly efficient and performant local control using gradient-based trajectory optimization and model predictive control. Finally, I remark on some fundamental limitations of local control and how we can connect the proposed local control capability with more global sampling-based motion planners to achieve long-horizon global plans. The proposed method achieves contact-rich plans such as dexterous in-hand reorientation and whole-body manipulation much more efficiently than RL while being highly scalable compared to methods that explicitly reason about contact modes. These results achieve a reduction of contact-rich manipulation to kinodynamic motion planning, and exposes the true difficulty of contact-rich manipulation from combinatorial explosion in contact modes to combinatorial and highly non-local decisions over motion planning behaviors. 32-D463 (Star)

October 25, 2024

Modern Clouds: Side-Channel Attacks and Defenses

Neil Zhao
NVIDIA Research and The University of Texas at Austin
2:00P
- 3:00P

Location

Seminar Room G882 (Hewlett Room)
Add to Calendar 2024-10-25 14:00:00 2024-10-25 15:00:00 America/New_York Modern Clouds: Side-Channel Attacks and Defenses Abstract:Cloud computing, which has seen significant growth over the past decade, fundamentally relies on the sharing of hardware resources among users. This approach enhances resource utilization and reduces operational costs. However, it also enables unintended information leakage through microarchitectural side channels. Despite the threat of side-channel attacks, cloud vendors remain skeptical about the practicality of these attacks in production cloud environments, leading to inadequate side-channel mitigations. My research focuses on exploring side-channel attacks in realistic cloud settings and developing comprehensive defenses across the computing stack. In this talk, I will first introduce a series of novel attack techniques that address practical challenges in conducting side-channel attacks in public clouds. Using these techniques, I demonstrated an end-to-end, cross-tenant side-channel attack on Google Cloud. This demonstration was subsequently recognized by Google as a critical-level bug, prompting a review by their product team. In the second part of this talk, I will introduce Untangle, a novel framework for side-channel defense. Untangle is designed to quantify and reduce information leakage in defense schemes based on dynamic resource-partitioning. Untangle opens up a defense paradigm that allows a controlled amount of information leakage in exchange for improved performance. To conclude, I will outline future research directions aimed at developing secure and efficient cloud systems resistant to side-channel attacks. Bio:Neil Zhao is currently a post-doc researcher at NVIDIA Research. Starting in the Fall of 2025, he will be an Assistant Professor in the Chandra Family Department of Electrical and Computer Engineering at The University of Texas at Austin. Neil received his PhD in Computer Science from the University of Illinois Urbana-Champaign (UIUC) in 2024, where he was advised by Prof. Josep Torrellas. His research interests include Computer Architecture, Hardware/System Security, and Cloud Computing. For more information about his work and achievements, please visit https://neilzhao.me. Seminar Room G882 (Hewlett Room)

November 01, 2024

Security Seminar

Mikko Hypponen, Ben Adida
WithSecure, VotingWorks

Part Of

2:00P
- 4:30P

Location

32-155
Add to Calendar 2024-11-01 14:00:00 2024-11-01 16:30:00 America/New_York Security Seminar Program:2:00pm - 3:00pmTECHNOLOGY REVOLUTIONSMikko Hypponen, WithSecureAbstract:All new technical innovations come with both advantages and disadvantages; we cannot simply select the benefits without also encountering the challenges. Once something is invented, we can't make it go away. This applies to things like artificial intelligence, the Tor Network, cryptocurrencies, strong encryption, quantum computing - even the internet itself.Mikko Hypponen is one of the most recognized cyber security experts world-wide, a keynote speaker and a best-selling author. Mikko works as the Chief Research Officer for WithSecure in Finland. He has served as an advisor for EUROPOL and the Monetary Authority of Singapore. Mikko's TED Talk has been seen by 2 million people, and his latest book has been translated to 5 languages.3:00pm - 3:30pm - Coffee Break3:30pm - 4:30pmA Voting Machine Everyone Can TrustBen Adida, VotingWorks.SB '98, MEng '99, PhD '06 Abstract:What does it take to build a voting machine that every voter can trust? Over the last 5 years, at VotingWorks, we've been working in the open on exactly that problem. We've concluded that the broadly accepted expert recommendations – paper ballots and post-election audits – are certainly necessary but far from sufficient. With public scrutiny into voting systems at an all-time high, and confidence in the outcome of our elections becoming increasingly partisan, we propose a new standard for voting machines – a standard that includes real transparency, strong system integrity, and a focus on simplicity. In this talk, we'll cover the why and the how, and we'll have voting equipment for attendees to try. 32-155

November 14, 2024

Cardinality Estimation using Lp norms on Degree Sequences

Dan Olteanu
University of Zurich
1:00P
- 2:00P

Location

Star - 32-D463
Add to Calendar 2024-11-14 13:00:00 2024-11-14 14:00:00 America/New_York Cardinality Estimation using Lp norms on Degree Sequences AbstractCardinality estimation is the problem of estimating the size of the output of a query, without actually evaluating the query. The cardinality estimator is a critical piece of any query optimizer, and is often the main culprit when the optimizer chooses a poor plan.In this talk, I will introduce LpBound, a pessimistic cardinality estimator that computes a guaranteed upper bound on the size of the query output. Besides common statistics used for cardinality estimation, such as the sizes of the input relations, the domain sizes for various columns, histograms, and the most common key values in join columns, LpBound also uses Lp norms on the degree sequences of the join columns. The computed cardinality bound is an optimal solution of a linear program whose constraints encode such statistics and Shannon inequalities.I will also report on an on-going experimental evaluation of LpBound for a variety of queries from the JOB, STATS, and subgraph matching benchmarks: acyclic and cyclic queries, full and group-by queries, and queries with and without range and equality predicates. LpBound takes time and uses memory very close to the traditional estimators used by PostgreSQL, DuckDB, and a commercial database system. Yet the range of LpBound overestimates is much smaller than the range of the estimates of the traditional estimators, which can both underestimate and overestimate by orders of magnitude. When fed the LpBound cardinalities, PostgreSQL can take less query evaluation time than when using its own estimates.This is joint work with Mahmoud Abo Khamis (RelationalAI), Chris Mayer (U Zurich), Dan Suciu (U Washington), and Haozhe Zhang (U Zurich).BioDan Olteanu is a professor at the University of Zurich, where he leads the Data Systems and Theory group (https://www.ifi.uzh.ch/en/dast.html), and a computer scientist at RelationalAI (https://relational.ai), where he works on incremental view maintenance, cardinality estimation, in-database machine learning, and data science projects.

December 01, 2024

Cancelled

Test event

This event has been cancelled

12:00P
- 1:00P

December 03, 2024

December 04, 2024

February 19, 2025