AI transformation happens when better intelligence becomes better decisions.
At Percepta, we’re using frontier AI to model operations, human behavior, and outcomes—and a modern take on operations research to turn those models into action. My co-founder @apjacob03 explains.
Constantinos Daskalakis
Constantinos Daskalakis (Greek: Κωνσταντίνος Δασκαλάκης; born 29 April 1981) is a Greek theoretical computer scientist. He is a professor at MIT's Electrical Engineering and Computer Science department and a member of the MIT Computer Science and Artificial Intelligence Laboratory. He was awarded the Rolf Nevanlinna Prize and the Grace Murray Hopper Award in 2018. More on Wikipedia
-
Constantinos Daskalakis
@KonstDaskalakis
Athul Paul Jacob
@apjacob03
-
Constantinos Daskalakis reposted
Sitan Chen
@sitanch
🚨 Thrilled to be co-organizing a Simons Institute semester-long program on diffusions and flows with @joanbruna, Sinho Chewi, @KonstDaskalakis, @ArnaudDoucet1 in Fall '27!
The call for Simons postdoctoral research fellowships for the program is out now, more info below ⬇️ 1/3 -
Constantinos Daskalakis reposted
Eugene Vinitsky 🦋
@EugeneVinitsky
I'm hiring for our interpretable world modeling effort at Percepta! The core question: how can we make an agent build a simulator, understand a company, and then design a policy, all from unstructured data? Come work with me and our amazing team.
jobs.ashbyhq.com/percepta/909e2fcb-98e3-4c6b-9b76-6c54091… -
Constantinos Daskalakis reposted
Ryan O'Donnell
@BooleanAnalysis
STOC 2027 Call for Papers is out:
acm-stoc.org/stoc2027/stoc2027-cfp.html
Note quite a lot of changes from previous STOC CfPs. -
Constantinos Daskalakis reposted
Aaron Roth
@Aaroth
A world without open problems
Here are some that fell today:
K-server: arxiv.org/abs/2609.15979
Matroid Secretary: arxiv.org/abs/2609.14555v1
Matrix Spencer: arxiv.org/abs/2609.15025
(Well Matrix Spencer was maybe also a few weeks ago, but who's counting? arxiv.org/abs/2608.28816 )
Aaron Roth
@Aaroth
We are clearly moving to a world that will be without open problems, but this doesn't mean math is going away. Interesting work will correspond to discovering new questions and coherent theoretical frameworks. Well defined agreed-to-be-interesting problems won't last long.
-
Constantinos Daskalakis reposted
MIT CSAIL
@MIT_CSAIL
MIT researchers developed a faster way to sample from AI diffusion models.
They propose a new algorithm reaching the same level of accuracy in exponentially fewer steps, suggesting a path to much more efficient image generation & other diffusion-based AI systems. It recently won Outstanding Paper at ICML: arxiv.org/abs/2602.01338 -
Constantinos Daskalakis
@KonstDaskalakis
When AI targets tasks in the physical world, it is often bound by the lack of large volumes of high-quality data. We've developed a general methodology for training diffusion models using a combo of high- & low-quality data. Applied to robotics, it gives sota diffusion policies👇
Adam Wei @CoRL ✈️
@adamwei_
🤖 We introduce Ambient Diffusion Policy, a simple and principled method for training policies with suboptimal data in robotics.
Suboptimal data is everywhere in robotics…
❌ Data filtering is wasteful
❌ Co-training learns both good and bad features
✅ Ambient Diffusion Policy selectively learns useful features via noise-dependent data usage
👇🧵(1/5)This media couldn't be displayed. See it on the original post.
Constantinos Daskalakis
@KonstDaskalakis
Applied to the generation of de novo protein structures, it gave sota protein diffusion models 👇
Giannis Daras
@giannis_daras
Announcing Ambient Protein Diffusion, a state-of-the-art 17M-params generative model for protein structures.
Diversity improves by 91% and designability by 26% over previous 200M SOTA model for long proteins.
The trick? Treat low pLDDT AlphaFold predictions as low-quality data
Constantinos Daskalakis
@KonstDaskalakis
The general, application-agnostic methodology is discussed here 👇
Giannis Daras
@giannis_daras
Announcing Ambient Diffusion Omni — a framework that uses synthetic, low-quality, and out-of-distribution data to improve diffusion models.
State-of-the-art ImageNet performance. A strong text-to-image results in just 2 days on 8 GPUs.
Filtering ❌
Clever data use ✅ -
Constantinos Daskalakis reposted
Adam Wei @CoRL ✈️
@adamwei_
🤖 We introduce Ambient Diffusion Policy, a simple and principled method for training policies with suboptimal data in robotics.
Suboptimal data is everywhere in robotics…
❌ Data filtering is wasteful
❌ Co-training learns both good and bad features
✅ Ambient Diffusion Policy selectively learns useful features via noise-dependent data usage
👇🧵(1/5)This media couldn't be displayed. See it on the original post. -
Constantinos Daskalakis
@KonstDaskalakis
Proud advisor moment! Congratulations @GolowichNoah!
ACM SIGecom
@AcmSIGecom
📢 Announcing the 2025 SIGecom Doctoral Dissertation Awardees!
🏆 Winner: @GolowichNoah (MIT), advised by @KonstDaskalakis and Ankur Moitra, for the thesis: "Theoretical Foundations for Learning in Games and Dynamic Environments" -
Constantinos Daskalakis reposted
Athul Paul Jacob
@apjacob03
1/N Over the past year, one thing that has become increasingly clear to me is that timelines are compressing very very quickly.
That changes what is worth attempting. The problems that would have sounded too early, or too ambitious even recently are starting to look tractable. -
Constantinos Daskalakis reposted
Athul Paul Jacob
@apjacob03
Heading to #neurips next week (Dec 2–5) with Percepta and @generalcatalyst! Would love to chat about auto-formalization, operations research, RL, foundation models, game theory, startups, and venture.
Drop me a DM if you’d like to meet! -
Constantinos Daskalakis reposted
Aaron Roth
@Aaroth
If you work at the intersection of CS and economics (or think your work is of interest to those who do!) consider submitting to the ESIF Economics and AI+ML meeting this summer at Cornell: www.econometricsociety.org/regional-activities/schedule/2…
-
Constantinos Daskalakis reposted
Athul Paul Jacob
@apjacob03
Today marks an important milestone.
I’m launching Percepta together with @htaneja, @hirshjain, @tmathew0309, Radha Jain, @marisbest2, @KonstDaskalakis and an incredible team, with the goal of bringing AI to the core industries that run our economy. For AI to deliver transformational impact, we need a strategy that orchestrates multiple disciplines. The most meaningful work of my career has been with strong interdisciplinary teams who take on ambitious challenges.
That’s the kind of team we’re building at Percepta.
Two pillars key pillars our work:
First, we deeply partner. Our researchers, product managers, and engineers work inside our partners’ operations - shadowing workflows, mapping decision points, building AI systems, and ultimately owning outcomes.
Second, we believe organizations revolve around decisions, and we’re reimagining decision-making to orchestrate AI agents and humans. This is our core research focus, blending foundation models, reinforcement learning, and optimization. The unique challenges of these domains won’t be solved by mimicking human data or finding a better prompt. Decision-making has many distinct elements, and our research team brings complementary skills to the challenge: collectively, we’ve built superhuman bots, deployed RL agents at scale, trained large language models to reason, proved fundamental results in learning theory, and released algorithms and tools used by the field every day.
We’ve already seen outcomes that change behavior: step-function productivity improvements in financial services, material enterprise value in care delivery organizations, and orders-of-magnitude cycle-time reductions in public-sector processes.
A bit over a year ago today, I joined @generalcatalyst - compelled by @htaneja vision to partner with him and help establish the early workings of Percepta. Since then, we’ve been quietly building a team of 25+.
We’re based in NYC and Boston and growing quickly. I’m grateful to the broader @generalcatalyst famiglia and especially @quentinclark, Jeannette zu Fürstenberg, @marcbhargava, David Fialkow, and @teresacarlson for being early partners in our journey.
We’re working against an incredibly ambitious mission. If this excites you, let’s chat!
Learn more:
www.percepta.ai
www.generalcatalyst.com/stories/unveiling-percepta -
Constantinos Daskalakis
@KonstDaskalakis
💪
Eugene Vinitsky 🦋
@EugeneVinitsky
Hiring researchers and engineers for a stealth, applied research company with a focus on RL x foundation models. Folks on the team already are leading RL / learning researchers. If you think you'd be good at the research needed to get things working in practice, email me
-
Constantinos Daskalakis reposted
Eugene Vinitsky 🦋
@EugeneVinitsky
Hiring researchers and engineers for a stealth, applied research company with a focus on RL x foundation models. Folks on the team already are leading RL / learning researchers. If you think you'd be good at the research needed to get things working in practice, email me
-
Constantinos Daskalakis reposted
Constantine Caramanis
@cmcaram
🚀 🇬🇷 A year in the making! I’ve just completed a set of 21 lectures in Machine Learning, in Greek, designed for high school students. The course introduces key ML concepts, coding in Python & PyTorch, and real-world AI applications. #machinelearning #ai #edtech #greece
-
Constantinos Daskalakis
@KonstDaskalakis
Proud of my brother's research!
SysBioStress - Daskalakis Lab
@NikosDaskalakis
Very proud and energized for 2025 that our paper on 🧠 molecular pathology in #ptsd and #mdd made it in #nih Director’s 2024 #sciencehighlights!
-
Constantinos Daskalakis
@KonstDaskalakis
Looking forward to this event on Thursday! www.kathimerini.gr/pages/announcements/563078821/aristote…
John Tasioulas
@JTasioulas
June 20th in Athens 🇬🇷- the Lyceum Project: AI Ethics with Aristotle. With speakers including @alondra, @KonstDaskalakis, @yuvalshany1, @kmitsotakis, Josiah Ober, @mbrendan1, and @FotiniChristia.
Places are rapidly filling up. Register now.
www.oxford-aiethics.ox.ac.uk/lyceum-project-ai-ethics-ari… -
Constantinos Daskalakis reposted
SysBioStress - Daskalakis Lab
@NikosDaskalakis
💡Daskalakis & Ressler labs @McLeanHospital @harvardmed are extremely excited & proud 🥳to share that our collaborative work with Nemeroff @UTAustin and Kleinman @LieberInstitute labs on molecular underpinnings of #ptsd & #mdd is out @ScienceMagazine www.science.org/doi/10.1126/science.adh3707! 🧵👇
-
Constantinos Daskalakis
@KonstDaskalakis
Can you train a generative model using only noisy data? If you can, this would alleviate the issue of training data memorization plaguing certain genAI models.
In exciting work with @giannis_daras and @AlexGDimakis we show how to do this for diffusion-based generative models.
Giannis Daras
@giannis_daras
Consistent Diffusion Meets Tweedie.
Our latest paper introduces an exact framework to train/finetune diffusion models like Stable Diffusion XL solely with noisy data.
A year's worth of work breakthrough in reducing memorization and its implications on copyright 🧵 -
Constantinos Daskalakis reposted
sxysun
@sxysun1
Stoked to announce an Agentic Markets workshop @agenticmarkets at #icml 2024! @icmlconf
📇 Details: sites.google.com/view/amw-2024/home
✏️ Call for papers: due May 17th
📅 Conference: July 26/27th
featuring GOAT speakers like Tuomas Sandholm, Gillian Hadfield, @drimgemp @KonstDaskalakis @NathanWorsley_ @StefLeonardos and more!
Are you ready for the internet of agents in high-stakes markets?
What economic paradigms will emerge when agents inevitably interact?
What kind of market should we design to "make incentives flow?"
Let's work on finding it out.
Meanwhile, some context and food for thought 👇 -
Constantinos Daskalakis reposted
John Tasioulas
@JTasioulas
The call for papers is coming soon… Delighted to confirm the participation of the Greek Prime Minister @kmitsotakis and the chair of Greece’s high level AI committee @KonstDaskalakis at this event. June 20th in Athens 🇬🇷
John Tasioulas
@JTasioulas
Plans are afoot to hold a workshop on Aristotelian approaches to AI ethics in Athens this summer (second half of June). This is an event I’m co-organising with Prof Josiah Ober and generously supported by @PJMFnd. We are especially keen on involving younger scholars. More soon.
-
Constantinos Daskalakis
@KonstDaskalakis
Exciting work w/ @YuvalDagan3 @MFishelson @GolowichNoah on efficient algos for no-swap regret learning and, relatedly, correlated eq when the #actions is exponentially large/infinite. While classical works point in the opposite direction, we show that this is actually possible!
Post not available here, view on X
Constantinos Daskalakis
@KonstDaskalakis
Check out our paper at: arxiv.org/abs/2310.19786.
Constantinos Daskalakis
@KonstDaskalakis
Note that concurrent work by Peng & Rubinstein establishes a similar set of results: arxiv.org/abs/2310.19647
-
Constantinos Daskalakis reposted
Alex Dimakis
@AlexGDimakis
Athens has a lot of talented AI researchers and developers.
orph
@orph_
Sequoia's new report on European technical talent lists Athens as one of the top 5 cities for AI
-
Constantinos Daskalakis reposted
Elias Papaioannou
@EliasPapaioann2
@DAcemogluMIT (@MITEcon) will present #powerandprogress (with @baselinescene) to the Greek audience at an online event organized @IOBE_FEIR on Wednesday, July 5 5.30pm Athens time @KonstDaskalakis (@MIT), Nikos Vettas, and I will discuss with Daron #ai, #technology, & #politics
IOBE
@IOBE_FEIR
🗓️ Wed Jul 5th, 2023
🕐 17:30-19:00 EEST
The @IOBE_FEIR invites you to an e-discussion with Daron Acemoglu, Institute Professor of Economics - MIT, following the publication of his new book “Power and Progress” (co-authored with S. Johnson).
youtube.com/live/AJH_tB4K-6E -
Constantinos Daskalakis reposted
Gautam Kamath
@thegautamkamath
ACM names the latest class of Fellows: congrats to a whole bunch of great theorists, including @boazbaraktcs, Sandy Irani, Rafael Pass, Samir Khuller, David Mount and especially my advisor @KonstDaskalakis!!
Also: my awesome colleague @lintool!
www.acm.org/media-center/2023/january/fellows-2022 -
Constantinos Daskalakis reposted
Giannis Daras
@giannis_daras
New ICML paper: Score-Guided Intermediate Layer Optimization (SGILO).
We train diffusion models on the latent space of StyleGAN and we show provable mixing of Langevin Dynamics for random generators.
Reconstructions for *extremely sparse* (<1%) measurements.
A thread🧵(1/N) -
Constantinos Daskalakis
@KonstDaskalakis
While coarse correlated equilibria (CCE) are very tractable in one-shot games, @GolowichNoah, @KaiqingZhang and I show that, surprisingly, stationary (Markov) CCE are intractable in stochastic games, and thus Multi-Agent Reinforcement Learning (MARL):
arxiv.org/abs/2204.03991Post not available here, view on X
Constantinos Daskalakis
@KonstDaskalakis
Our learning algorithm improves on prior work whose time depends exponentially on the number of agents (known as the "curse of multi-agents") as well as V-learning, which beats the curse of multi-agents but computes non-Markovian CCE.
Constantinos Daskalakis
@KonstDaskalakis
Contemporaneous work by Jin, Muthukumar and Sidford shows a related lower bound:
arxiv.org/abs/2204.04186 -
Constantinos Daskalakis
@KonstDaskalakis
How fast does player regret grow in multi-player games? Standard no-regret learners give regret growing as √T in T rounds of interaction. Recent work w/ @MFishelson & @GolowichNoah brings this down to a near-optimal poly(log T)!
arxiv.org/abs/2108.06924Post not available here, view on X
Constantinos Daskalakis
@KonstDaskalakis
In follow-up work, joining forces with Ioannis Anagnostides, @gabrfarina and Tuomas Sandholm, we obtain similar bounds for no-internal regret, no-swap regret, and convergence to correlated equilibrium.
arxiv.org/abs/2111.06008
Constantinos Daskalakis
@KonstDaskalakis
Even more recent work by Georgios Piliouras, Ryann Sim and Stratis Skoulakis provides an implicit method with constant regret (w.r.t. T) albeit with polynomial dependence on the dimension.
arxiv.org/abs/2111.14737 -
Constantinos Daskalakis reposted
Rodney Brooks
@rodneyabrooks
It's a deal. August 19, 2031. My prediction is that this project will not register at all in any history of the previous ten years. So we'll understand that it was a totally insignificant announcement.
-
Constantinos Daskalakis reposted
Angeliki Koutsoukou-Argyraki
@AngelikiKoutso1
An amazing discovery today during my random stroll (someplace in Greece, in a secret location)! As I took a shortcut through a tiny street, I was thrilled to see that someone had filled the place with beautifully drawn Geometry theorems! [1/9]
-
Constantinos Daskalakis
@KonstDaskalakis
A great line-up for the July 13-14 FODSI workshop on ML for Algos! @AlexGDimakis-Yonina Eldar-@annadgoldie-@HeckelReinhard-Stephanie Jegelka-@tim_kraska-Benjamin Moseley-David Parkes-@AlgoSvensson-Tuomas Sandholm-@vsergei-Ellen Vitercik-David Woodruff!
fodsi.us/ml4a.html
Constantinos Daskalakis
@KonstDaskalakis
The question: How to use ML+data to design algorithms w/ better performance on non worst-case instances, e.g. better data structures, online algos, streaming and sketching algos, market mechanisms and algos for combinatorial optimization, similarity search and inverse problems?
Constantinos Daskalakis
@KonstDaskalakis
Brought to you by Foundations of Data Science Institute: fodsi.us
-
Constantinos Daskalakis reposted
Market Design Community
@econD47
6⃣ talks on the agenda this week: @JacksonmMatt in @VMDSeminar, Marina Halac in Seminars in Economic Theory, @ArielRubinstein in @CaltechEconThry, Konrad Mierendorff in @VSET_contact, @KonstDaskalakis in Zurich Market Design (von Neumann Lecture), and Na Li in @Games_Seminar 👇👇
-
Constantinos Daskalakis reposted
Gautam Kamath
@thegautamkamath
Check out our Research Highlight in Communications of the ACM (@CACMmag)! "Robustness Meets Algorithms," on our work which eased tensions between computation and robustness in multivariate estimation. With Diakonikolas, Kane, @jerryzli, Moitra, Stewart.
cacm.acm.org/magazines/2021/5/252181-robustness-meets-alg… -
Constantinos Daskalakis reposted
Silvio Micali
@silviomicali
.@Algorand was designed to be environmentally friendly. 🌱
-
Constantinos Daskalakis
@KonstDaskalakis
Many thanks to Kush Bhatia and @CyrusRashtchian for their excellent article on this talk.
Gautam Kamath
@thegautamkamath
4/n Next up, we have an interview with @KonstDaskalakis and coverage of his keynote talk(www.youtube.com/watch?v=GpaCWKlOMig). Written by Kush Bhatia and @CyrusRashtchian. Thanks a lot to @boazbaraktcs for hosting this one! windowsontheory.org/2021/04/30/alt-highlights-equilibrium…
-
Constantinos Daskalakis
@KonstDaskalakis
Congratulations to Manolis Zampetakis for this richly deserved honor!
ACM SIGecom
@AcmSIGecom
The @AcmSIGecom Dissertation Award for 2020 goes to
Manolis Zampetakis for his thesis "Statistics in High Dimensions without IID Samples: Truncated Statistics and Minimax Optimization" advised by @KonstDaskalakis at @MIT. Congratulations!
More details: sigecom.org/award-phd.html
Constantinos Daskalakis
@KonstDaskalakis
In his beautiful thesis he expands Econ-CS research in fundamental & timely directions, pertaining to statistical estimation, optimization & learning from (i) incomplete (ie biased, censored, or truncated) data & (ii) data that is subject to adversarial or strategic manipulation.
Constantinos Daskalakis
@KonstDaskalakis
Many congratulations to the other awardees as well!
-
Constantinos Daskalakis
@KonstDaskalakis
How to use subadditivity properties of probability divergences in GANs to decrease the effective dimensionality of your problem?
Check out our oral paper with @FeiziSoheil and M. Ding at AISTATS in a couple of hours!
Soheil Feizi
@FeiziSoheil
Q: how can we improve the design of GANs if the underlying independence graph of variables (as a Bayes net or an MRF) is known?
A: check out our paper: arxiv.org/abs/2003.00652
which will be presented as an oral talk @aistats_conf
Joint work with M. Ding and @KonstDaskalakis -
Constantinos Daskalakis reposted
Games, Decisions & Networks Seminar Series
@Games_Seminar
Super excited to announce a new online seminar series on “Games, Decisions, and Networks”, aiming at bringing together researchers working on foundations and applications of Game, Decision, and Network Theory from computer science, control, economics, and operations research.
-
Constantinos Daskalakis
@KonstDaskalakis
Samples from high-dimensional distributions can be scarce or expensive to acquire. Can we meaningfully learn them from *one* sample?!? In new work w/ @YuvalDagan3, @NishanthDikkala, &Vardis Kandiros, we show how to learn Ising models given a single sample
arxiv.org/abs/2004.09370
Constantinos Daskalakis
@KonstDaskalakis
We quantify the error in estimating the model w.r.t. the metric entropy of possible interaction matrices. If multiple samples are available, we handle them by viewing them as one sample from a larger model. We thus provide one method for learning given one, a few, or many samples
Constantinos Daskalakis
@KonstDaskalakis
Our guarantees unify, extend and interpolate between two separate strands of works: recent work in CS on estimating MRFs from multiple independent samples under mild assumptions; and a growing literature in Probability on estimating them from one sample under more assumptions.
Constantinos Daskalakis
@KonstDaskalakis
On the technical front, we establish novel concentration and anti-concentration bounds for functions of the Ising model, a fascinating topic on its own right. We go beyond high temperatures! Learn more by attending a Simons workshop tomorrow@
simons.berkeley.edu/workshops/hd-2020-3 -
Constantinos Daskalakis
@KonstDaskalakis
Standard no-regret learners, e.g. online GD, converge to equilibria of monotone games like the moon converges to the earth, i.e. their average converges, at a best achievable rate of 1/T. At what rate can their last-iterate converge? See 👇 & find @GolowichNoah at NeurIPS today.
Noah Golowich
@GolowichNoah
New paper with Sarath Pattathil & Costis Daskalakis: arxiv.org/abs/2010.13724.
We answer the question: at what rate can players' actions converge to equilibrium if each plays according a no-regret algorithm in a smooth monotone game? -
Constantinos Daskalakis
@KonstDaskalakis
There are many algorithms for learning Ising models. Yet there is no sample-efficient and time-efficient one that works without assumptions, even for tree-structured models. We obtain the first such result in fresh work w/ @PanQinxuan:
arxiv.org/abs/2010.14864Post not available here, view on X
Constantinos Daskalakis
@KonstDaskalakis
Constantinos Daskalakis
@KonstDaskalakis
To control the error accumulation, we combine this hierarchical classification with delicate uses of the subadditivity of the squared Hellinger distance over graphical models from:
arxiv.org/abs/1612.03164 -
Constantinos Daskalakis
@KonstDaskalakis
Min-max optimization (used among other applications in GANs, and adversarial training more broadly) is empirically challenging.
We show why min-max optimization is hard in the following paper with Stratis Skoulakis and Manolis Zampetakis:
arxiv.org/abs/2009.09623Post not available here, view on X
Constantinos Daskalakis
@KonstDaskalakis
Said differently, computing approximate local min-max equilibria of two-player zero-sum games with smooth and Lipschitz objectives is PPAD-complete, i.e. as hard as (i) computing Nash equilibria in general-sum normal-form games; and (ii) computing approximate Brouwer fixed points
Constantinos Daskalakis
@KonstDaskalakis
Thus, while min-max equilibria of two-player zero-sum games with convex-concave objectives are tractable, even approximate and local min-max equilibria of nonconvex-nonconcave objectives are as intractable as Nash equilibria in general-sum normal-form games.
-
Constantinos Daskalakis
@KonstDaskalakis
Congratulations Manoli! Proud for the beautiful work in this dissertation!
Gautam Kamath
@thegautamkamath
Congratulations to my academic brother Manolis Zampetakis on successfully defending his PhD thesis! The latest amazing student of @KonstDaskalakis. Huge turnout for his defense (larger than expected). Going to UC Berkeley for a postdoc next. Check him out: mit.edu/~mzampet/
-
Constantinos Daskalakis reposted
Alex Dimakis
@AlexGDimakis
(1/2) For the first time in history (since its original showing in 472 BC) a drama (The Persians by Aeschylus) will be live-streamed from the magnificent Ancient Theater of Epidaurus.
www.thenationalherald.com/culture/arthro/epidaurus_ancien… -
Constantinos Daskalakis
@KonstDaskalakis
Congratulations to my student, Dr. @NishanthDikkala, who successfully defended today, having done innovative and deep work on statistical inference using dependent data!
Gautam Kamath
@thegautamkamath
Congratulations to my academic brother Dr. Nishanth Dikkala (@NishanthDikkala) who just defended his thesis on statistical inference from dependent data! Another student lucky to be advised by Costis (@KonstDaskalakis)! Check out his work on his website: people.csail.mit.edu/nishanthd/
-
Constantinos Daskalakis reposted
Alex Dimakis
@AlexGDimakis
People in the US need to practice aggressive social distancing before this happens
Jason Van Schoor
@jasonvanschoor
From a well respected friend and intensivist/A&E consultant who is currently in northern Italy:
1/ ‘I feel the pressure to give you a quick personal update about what is happening in Italy, and also give some quick direct advice about what you should do. -
Constantinos Daskalakis
@KonstDaskalakis
Identity testing between high-dimensional distributions requires exponentially many samples in the dimension---aka good luck GANs targeting high-dimensional distributions. Joint work with @FeiziSoheil and M. Ding shows how to escape these lower bounds if the target is a Bayesnet!
Soheil Feizi
@FeiziSoheil
Q: How to design a GAN for distributions on Bayesian networks? A: check out our paper introducing "SubAdditive GANs": arxiv.org/abs/2003.00652
Joint work with Constantinos Daskalakis @KonstDaskalakis and @umdcs student Mucong Ding, started at @SimonsInstitute -
Constantinos Daskalakis reposted
Lionel Page
@page_eco
One of my favourite examples of how people react to economic incentives:
Architectural tax avoidance👇
🇬🇧 UK : tax on windows
🇻🇳 Vietnam: tax on frontage
🇫🇷 France: tax on floors (roof exempted)
🇧🇷 Brazil: tax on church construction (when finished) -
Constantinos Daskalakis
@KonstDaskalakis
Happy advisor moment! Congratulations to Matt Weinberg and the other fellows!
Gautam Kamath
@thegautamkamath
The @SloanFoundation fellows for 2020! Non-exhaustive list of some names I recognize in CS&Math: @mcarbin, Sanjam Garg, @zicokolter, Yin Tat Lee, @polikarn , Aaron Sidford, Matt Weinberg, Hao Huang, @weijie444, and Cynthia Vinzant. Congrats to all winners! sloan.org/fellowships/2020-Fellows
-
Constantinos Daskalakis reposted
Gautam Kamath
@thegautamkamath
My PhD advisor @KonstDaskalakis makes his second appearance in an @NPR @planetmoney podcast, this time on how to allocate slots to the @nycmarathon.
www.npr.org/transcripts/793488868 -
Constantinos Daskalakis
@KonstDaskalakis
Honored to give a lecture in memory of Paris Kanellakis!
Brown CS
@BrownCSDept
Professor @KonstDaskalakis of @MIT will deliver the 19th annual Paris C. Kanellakis Memorial Lecture ("Learning from Censored and Dependent Data") at 4 PM on December 5 in CIT 368. Details at: cs.brown.edu/events/kanellakis
-
Constantinos Daskalakis reposted
Lionel Page
@page_eco
Such a cool way to explain logic gates.
-
Constantinos Daskalakis
@KonstDaskalakis
#papafest panel on the future of AI!
-
Constantinos Daskalakis
@KonstDaskalakis
#papafest starting now!
-
Constantinos Daskalakis
@KonstDaskalakis
Great lineup of invited speakers at the NeurIPS workshop on the important interface between ML and Game Theory! Submit your interesting papers!
Ioannis Mitliagkas (Γιάννης Μητλιάγκας)
@bouzoukipunks
Game theory and deep learning workshop at NeurIPS!
sgo-workshop.github.io
Invited talks:
Éva Tardos! Asu Ozdaglar! David Balduzzi! Fei Fang!
Plus more contributed talks, posters, panels and discussions.
Submit your extended abstracts by September 16th! -
Constantinos Daskalakis reposted
Tim Roughgarden
@Tim_Roughgarden
Conference celebrating Christos Papadimitriou (September 6-8, 2019, at Columbia University) papafest.cs.columbia.edu
-
Constantinos Daskalakis
@KonstDaskalakis
Congratulations to the ACM Turing Award winners-live from San Francisco
-
Constantinos Daskalakis reposted
Alex Dimakis
@AlexGDimakis
Updated Robust Manifold Defense paper
arxiv.org/abs/1712.09196
with
@andrew_ilyas @KonstDaskalakis.
We solve an open problem by @anishathalye on fully defeating DefenseGAN
and obtain SOTA MNIST robustness @aleks_madry
Code: github.com/ajiljalal/manifold-defense -
Constantinos Daskalakis
@KonstDaskalakis
Nice!
Gautam Kamath
@thegautamkamath
Adding airline pilot to my CV
-
Constantinos Daskalakis reposted
Gautam Kamath
@thegautamkamath
A big step for the open-access movement: the @UofCalifornia system discontinues their subscription to @ElsevierConnect. Will other universities follow suit?
Meredith Turner
@UC_Newsroom
“I fully support our faculty, staff and students in breaking down paywalls that hinder the sharing of groundbreaking research"—UC President Janet Napolitano on #openaccess publishing. www.universityofcalifornia.edu/press-room/uc-terminates-s…
-
Constantinos Daskalakis
@KonstDaskalakis
Proud advisor moment: congratulations @Yang_Cai for this well-deserved honor!
-
Constantinos Daskalakis reposted
Jordan Ellenberg
@JSEllenberg
A zero-sum game is, by definition, one where not everyone can lose!
www.wired.com/story/new-york-and-amazon-played-a-zero-sum… -
Constantinos Daskalakis reposted
Nicholas A. Christakis
@NAChristakis
An MD cracked his knuckles for 50 years, but only for one hand, and concludes that both the claim that cracking causes arthritis, and that parents are right, must be rejected. Fabulous N-of-1 trial with internal control. onlinelibrary.wiley.com/doi/pdf/10.1002/1529-0131(199805)… Via @EikoFried @WeedenKim
-
Constantinos Daskalakis
@KonstDaskalakis
Just posted a photo www.instagram.com/p/Bs1LmLqHml6/
-
Constantinos Daskalakis reposted
Gautam Kamath
@thegautamkamath
An obituary for Jean Bourgain from @nytimes, mentions a good research habit: "[He] (...) kept two problems in mind: 'One which I want to solve and another which I can solve.'" When stuck on the former, he would work on the latter as a confidence boost.
www.nytimes.com/2019/01/16/obituaries/jean-bourgain-dead.… -
Constantinos Daskalakis reposted
Timothy Gowers @wtgowers
@wtgowers
Oh dear, yet another mathematical giant has left us (after Peter Swinnerton-Dyer, Elias Stein and Jean Bourgain). He was also a former master of Trinity College Cambridge, and extraordinarily energetic right into his late 80s.
N. Ghoussoub
@NGhoussoub
Sorry to confirm that Sir Michael Atiyah has died today. It is hard to believe that such a force of nature, life, and mathematical creativity is gone. RIP.
The Royal Society
@royalsociety
Michael Atiyah OM FRS, President of the Royal Society 1990 -1995, died today. He was "a wonderful person who was a true internationalist and a fervent supporter for investing in talent – themes which resonate very clearly today." Read the full tribute royalsociety.org/news/2019/01/tribute-to-former-president…
-
Constantinos Daskalakis reposted
Rodney Brooks
@rodneyabrooks
On this day in 1903 was born John von Neumann, the greatest polymath of the 20th century. We are all using "von Neumann architecture" chips to read this tweet. Game theory, ergodic theory, Artificial Life, quantum mechanics, hydrodynamics, statistics, etc. He was an immigrant.
-
Constantinos Daskalakis reposted
The New York Times
@nytimes
50 years after he debuted "The Art of Computer Programming," Donald Knuth reflects on his opus-in-progress. “It started out that computer scientists were worried nobody was listening to us. Now I’m worried that too many people are listening.” www.nytimes.com/2018/12/17/science/donald-knuth-computers…
-
Constantinos Daskalakis reposted
Alex Dimakis
@AlexGDimakis
www.nature.com/articles/d41586-018-07730-1
Time for European companies to put some more R into their R&D spending. -
Constantinos Daskalakis reposted
Ioannis Mitliagkas (Γιάννης Μητλιάγκας)
@bouzoukipunks
A couple of days left to apply for the research MSc and PhD programs at @MILAMontreal
We’re looking for students excited about research in ML, optimization, deep learning, NLP, RL, applications…
Hear the calling of 🇨🇦 and come join us :)
mila.quebec/en/admission/ -
Constantinos Daskalakis reposted
Paul Krugman
@paulkrugman
Well, Congressman, PageRank finds the eigenvector with the largest eigenvalue ... oh, never mind
Will Sommer
@willsommer
Rep. Steve King is demanding to know the names of Google search employees so he can go through their social media profiles to find out if they're liberals.
-
Constantinos Daskalakis reposted
Andrej Karpathy
@karpathy
Incredible NeurIPS talk from @drmichaellevin on "Bioelectric Computation Outside the Nervous System" www.youtube.com/watch?v=RjD1aLm4Thg so much focus on the software of neural networks when cellular networks (more generally) feature more ancient and highly sophisticated / robust dynamics
-
Constantinos Daskalakis reposted
Rodney Brooks
@rodneyabrooks
Driverless taxis with human drivers but no passengers. Reality is way harder than researchers think after their first few demos. First driverless vehicles were on freeways in 1987. It always takes longer than everyone thinks. www.azcentral.com/story/money/business/tech/2018/12/05/ph…
-
Constantinos Daskalakis reposted
Alex Dimakis
@AlexGDimakis
Yo-Yo Ma at NeurIPS!
-
Constantinos Daskalakis reposted
Alex Dimakis
@AlexGDimakis
CIFAR10
Reza Zadeh
@Reza_Zadeh
Major exports of Canada:
1) Maple Syrup
2) Deep Learning
3) Reinforcement Learning
4) ???
#neurips2018 -
Constantinos Daskalakis reposted
MIT IDSS
@mitidss
Congrats to Dimitri Bertsekas and John Tsitsiklis for being awarded the 2018 INFORMS John von Neumann Theory Prize for their fundamental, sustained contributions to theory in operations research and the management sciences. lids.mit.edu/news-and-events/news/dimitri-bertsekas-and-j…
-
Constantinos Daskalakis reposted
Arnab Bhattacharyya
@abhatt2
We are presenting our work on learning and testing causal models at poster #8, room 210 this morning for #neurips2018. Drop by if you're interested in causal inference or distribution learning and testing. #neurips



































