Shirley Ho
@cosmo_shirley
Followers
4K
Following
2K
Media
146
Statuses
952
Group Leader @FlatironCCA @SimonsFdn π Professor at @NYUPhysics @NYUDataScience @Princeton ML/AI-accelerated Science, @PolymathicAI HongKonger
Joined November 2014
I am so proud and excited to announce that @MilesCranmer successfully defended his Ph.D. thesis titled "Interpretable Machine Learning for the Physical Sciences" yesterday from Princeton University. #veryproudadvisor π₯°. HUGE thanks to @DavidSpergel for co-advising Miles w/me!
19
19
366
Let's play a game: . One of the rows here shows the formation of the Universe simulated by the computer painstakingly with the laws of physics, the other one is learned by AI. Top or Bottom is generated by AI? Please answer below π. This work led by @sciencedrew is pushing
31
57
311
We are hiring software-focused researchers at @PolymathicAI ! . Our goal is to develop large foundation models for the sciences.
7
40
275
Thanks to @ykilcher for explaining our paper with a youtube video! Paper led by @MilesCranmer w/ @PeterWBattaglia @KyleCranmer @DavidSpergel .
New Video π₯ Deep Learning is very good at fitting functions numerically, but what about deriving symbolic expressions? How Graph Networks can learn Newtonian Physics and Dark Matter!.@MilesCranmer @PeterWBattaglia @KyleCranmer @DavidSpergel @cosmo_shirley
1
55
266
Very happy to start this initiative @PolymathicAI with our amazing team members and scientific advisors!. The mission: Building #opensource AI models trained across disciplines that deliver scientific discoveries πͺ. Site: Github:
12
22
208
Being a first-gen college grad whose parents could not have afforded to see my graduations, it was really nice to be able to finally make them proud :) . Special thanks to @DavidSpergel, and @dalcantonJD for being the best supporters! . @BlavatnikAwards #ThankYOU!
18
9
202
The deadline to apply is in 9 days! Don't miss the opportunity to spend the coming summer in NYC fully supported to learn ML X Science! . @SimonsFdn's @FlatironInst will host the first #MachineLearning X #Science summer school .
6
83
213
Our internship program at Polymathic is open for opportunities from now through fall 2025! . I believe our program provides an opportunity to work alongside some of the best researchers and engineering experts in the world β exploring the unknown of building foundation models for
9
45
210
Have you all heard about ChatGPT or foundation models but want to build more than a chatbot with AI? . π₯ We at @PolymathicAI are building foundation models for science π₯ . Join us (@albertobietti @MilesCranmer @kchonyc @meickenberg @SiavashGolkar Francois Lanusse,
7
41
190
Are you a student looking for an internship or someone looking for a full-time software engineering position to possibly upend how scientific machine learning is done? . Join us at @PolymathicAI in NYC at @FlatironInst!. Reviewing starts in early Jan!
6
34
193
You heard all about AI accelerating simulations (maybe from me?), but do you know. How can AI tell you what is in the Universe?. Our new series of work from #SimBIG team led by @changhoon_hahn published recently by @NatureAstronomy did just that! . Interesting things we did:
Excited to announce that our latest #SimBIG collaboration research has just been published in @NatureAstronomy πβ¨!.#Astronomy #Cosmology #NatureAstronomy
6
53
202
Happy to announce that we will have a ML X Physics conference "AI-driven discovery in physics and astrophysics" at University of Tokyo's new Center for Data Driven Discovery on Jan 22-26, 2024 ! . Abstract Submission Deadline Nov 30, 2023!
2
33
157
Very excited about our recent paper on using machine learning to estimate the Galactic accelerations! . Led by @JakeNibauer from @Princeton
1
26
154
Want to build powerful AI for Science? . We at @PolymathicAI are building foundation models for science π₯ at @FlatironInst . Join us ( @albertobietti @MilesCranmer @kchonyc @meickenberg @SiavashGolkar @marielpettee, Francois Lanusse, @mikemccabe210, Tom Hehir, Rudy Morel,
3
23
128
I am part of a group of Hong Kong Americans who found 0.5 million masks for medical workers in the heavily hit areas: NY, NJ, CA, WA. But we can only distribute as many as we can pay for them! .Send a mask to medical workers:
We have successfully distributed some of our masks to the Mount Sinai Hospital (New York City) and Elmhurst hospital in Queens!. Please continue to support us:. #Hongkongers4US
8
49
108
Using #normalizingflows for #GravitationalWaves population analysis! [Left]: using normalizing flow for the likelihood. [Right]: using analytic likelihood. But you can make normalizing flow likelihood from training data w/o analytical assumptions
3
26
106
People often ask if AI will take over the Universe π€ͺ. That I don't know, but how about. Could AI discover what is in the Universe? . With our new approach, the answer is YES! . Led by @changhoon_hahn . Papers: .and .
6
20
104
Very proud of our team :) . @MilesCranmer @kchonyc @meickenberg Francois Lanusse, @marielpettee @albertobietti @leopoldosarra.@SiavashGolkar @mikemccabe210 @liamparker @liamhparker @oharub @BrunoRegaldo @khirashimaAstro
Very excited to share that our team's "Multiple Physics Pre-training" paper won the Best Paper Award at @AI_for_Science workshop at NeurIPS this year! . Congrats to the team and to @mikemccabe210, @BrunoRegaldo @liamhparker and @oharub for leading the effort! π₯³.
4
8
102
@SimonsFdn's @FlatironInst will host the first #MachineLearning X #Science summer school!.It is 8 weeks long with lectures and mentored research. We have an excellent lineup of lecturers: NoamBrown @gppcarleo @ylecun Barbara Engelhardt @kchonyc.
5
28
99
I am very proud of our first crew of #ML X #Science summer school students "graduating" yesterday at @FlatironInst . Special thanks to all mentors, lecturers for making this happen!! π @MilesCranmer @kchonyc @ylecun @polynoamial @gppcarleo Stephane Mallat, Barbara Engelhardt
2
11
95
Are you a Ph.D. student?. Do you want to come and work with us at Flatiron for a semester at the intersection of #astrophysics and #DeepLearning?. Check out the projects I posted here at on #transformer, #simulations & #astrophysics!
5
18
87
We are organizing "Cosmic Connections": .An AI X Astro Symposium in @SimonsFdn at @FlatironCCA! . If you are a student/researcher/faculty who is interested in AI X Astro area, consider applying to join us at the symposium. Apply by April 9! . Awesome speakers lineup π
4
19
85
Come here our selected contributed talk on simulating turbulence with deep learning at #SimDL ( today. Led by @neuro_kim, Alvaro Sanchez, @DrumBushField, Dmitrii Kochkov, @MilesCranmer, @spectralhippo, Jonathan Goodwin, Elaine Cui and @PeterWBattaglia
1
16
79
What percentage of academics will now use #ChatGPT to write introductions to their papers? π
.
21
7
71
π₯@PolymathicAI team will present Multiple Physics Pretraining (MPP) tomorrow from 11am to 2pm (PST) at East Exhibit Hall A-C #4100 π₯ at #NeurIPS2024! . Learn how to pre-train on various incompressible fluids simulations and make the AI model predict what will happen to nearly
3
10
78
Super pumped for our Machine Learning for Physical Sciences #neurips2020!. Remember to submit to our workshop #ML4PS!.
I'm thrilled that our Machine Learning for Physical Sciences #NeurIPS2020 workshop proposal was accepted! Last year we had a great turnout and fantastic talks - check out the #ML4PS hashtag. New to team: @adjiboussodieng @iamstarnord @glouppe @zdeborova
1
13
71
Excited to be giving an invited talk at Foundation Models for Science workshop at Neurips in ~1.5 hours about @PolymathicAI ! . Come join us at West Meeting Room 202-204! at 11:15am! π₯. Thank you to the organizers for making this workshop possible! .
5
5
71
π₯We are hiring! π₯. 7 more days to apply to work at amazing .@FlatironCCA with people like @davidwhogg @dalcantonJD @exoplaneteer @meg_bedell @jcolinhill @BlakesleyB @DrumBushField @physicskaze @DavidSpergel just to name a few! . Deadline October 15!.
1
26
70
Want excellent food, .street performance, music scene and .science at the same time?! . Come work with us at @FlatironCCA at @SimonsFdn in NYC! . We are hiring ! π₯. Deadline to apply Nov 1 . Postdoc: Software Postdocs: 1/6
4
19
71
Submit your paper to our awesome @NeurIPSConf workshop!!.
Excited to be organizing the NeurIPS workshop on Interpretable Inductive Biases and Physically Structured Learning!. Check it out and consider submitting a 4-page workshop paper - extended abstract deadline is October 2nd. @_mlutter @cosmo_shirley @wangleiphy
0
13
70
A bit of a fan girl moment yesterday @NASAGoddard when I saw the space environment simulator where @NASAHubble , #COBE, #WMAP, @NASAWebb had been shaken up, vibrated, and tested!
2
3
65
This is all thanks to my team, collaborators and many advisors+colleagues+friends along the way π».@DavidSpergel @MilesCranmer @changhoon_hahn, Joanne Cohn, Jill Knapp, Jim Gunn, Uros Seljak, Martin White, @PeterWBattaglia, @kchonyc, @AstroCKragh, @LoverdeMarilena @kheegster.
Congratulations to @cosmo_shirley on being named a finalist for the 2023 @BlavatnikAwards! Shirley Ho leads our Cosmology X Data Science group at @FlatironCCA and is a professor at @nyuniversity. Read more here:
12
9
64
#Shameless promotion of our #NeurIPS2020 #ML4PS2020 papers: . @astroVAV teaches us how to find weird #Supernova from time-series data.
1
8
61
You probably all heard about #ChatGPT, but want to build more than a chatbot with AI?. π₯ We at @PolymathicAI are building foundation models for science and you can be part of this!. Join us this summer or this fall as our intern at @FlatironInst in #NYC . Google form π app:
3
17
61
Excited to be at @NeurIPSConf along with the team @PolymathicAI starting Wednesday! . Find us π . @MilesCranmer @kchonyc @meickenberg Francois Lanusse, @marielpettee @albertobietti @leopoldosarra @SiavashGolkar @mikemccabe210 @liamparker @liamhparker @oharub @BrunoRegaldo.
We're thrilled to be at @NeurIPSConf for the first time since we formed a few months ago! . If you're there, we'd love to chat with you about our team and AI/ML research! . Reach out and be sure to check out our accepted papersπ#NeurIPS2023 #AI4Science. and we are HIRING π₯³
6
8
59
π€© Excited to see the first workshop on Foundation Models for Science at #NeurIPS2024!. Very much looking forward to speaking alongside excellent scientists Michael Mahoney, @ParisPerdikaris, Danielle Maddix Robinson, @wellingmax, and @laurezanna !!. Website:.
Working on foundation models for scientific problems? Consider submitting your paper to the 1st Workshop on Foundation Models for Science (FM4Science) #neurips2024 in Vancouver!. Abstract DDL: Aug. 27, 2024 AOE.OpenReview: Workshop:
0
7
60
Excited to be talking at the American Physical Society April meeting today in 40mins. Invited talk in session "Generative Models in Fundamental Physics". π Come watch it if you can! .I believe all talks will be available online later as well. #aps22.
3
2
56
Join us at @PolymathicAI at @SimonsFdn's @FlatironInst to create AI models for science! . Opportunities available for postdoctoral candidates! . Google form below: .
Want to do research in ML for science in NYC and develop the next generation of foundation models for scientific data? Apply here to be a postdoc at @PolymathicAI @SimonsFdn :
4
20
53
Excited to share our new publication from @PolymathicAI on @RAS_Journals: . AstroCLIP.The First cross-modal foundation model for galaxies! . Work led by @liamhparker, Francois Lanusse, @SiavashGolkar, @LeopoldoSarra and @MilesCranmer
The new-and-improved AstroCLIP has been published with the Monthly Notices of the Royal Astronomical Society!. AstroCLIP is the first cross-modal foundation model for galaxies, and competitively performs a wide array of astronomical tasks. Web App:
2
10
51
π₯ Shameless self-promotion here: Come hear this awesome ;) line up at "AI for Science: Mind the Gaps" #NeurIPS2021 workshop tomorrow, 8:30am-6pm EST! π₯.
1
8
46
Cosmic Connections symposium kicked off today with insightful remarks by @DavidSpergel at @SimonsFdn, Ashley Villar, and @kchonyc plenary talks!
4
6
46
Staring into the Voids . [not philosophy tweet, don't worry π€] . featuring our group's Alice Pisani, @bwandelt, @DavidSpergel at @FlatironCCA by @sciam! .
5
9
43
What does a #neuralnetwork learn when it has looked at tens of thousands of #Universes? . Our New @PNASNews (Proceedings of National Academy of Sciences) paper Led by Drew Jamieson shows what it had learned π₯ see thread below π. 1/N
1
7
43
Very proud of what we accomplished with @PolymathicAI + MultiModal Universe Collaboration! . π100 TBs of Astronomical data βοΈππ«.πFrom > 10 telescopes, over 20 modalities.πAll ML training ready . This will be presented today (Wed) at 4:30pm-7:30pm (PST) at West Ballroom A-D
0
9
45
What happens when you throw the #diffusion model (think #Dalle #StableDiffusion) into the #Universe? . We can *maybe* figure out what happened at the beginning of the Universe with error bars π₯. .Led by Ronan Legin (University of Montreal)
4
14
39
Excited to see this released and even happier to see that Cantonese is among the 100 languages being translated!.
Seamless4MT: Massive Multilingual Multimodal Machine Translation. Language translation + speech recognition + speech synthesis in a single model: speech-to-speech, text-to-text, speech-to-text and text-to-speech. Works for 100 languages. Code available under CC-BY-NC license.
1
2
38
One can imagine lots of physical applications of this adversarial latent autoencoder. But it would be interesting to morph our politicians into #GameofThrones characters . who should @realDonaldTrump be?.
Adversarial Latent Autoencoders. Kind of creepy to imagine what my entire family might look like as Tyrion Lannister, Neo, or Emma Watson.
2
7
39
Do you ever imagine what happened at the beginning of the Universe? . Our paper led by Vaibhav Jindal (@CarnegieMellon ), Drew Jamieson (MPA), Albert Liang (@CarnegieMellon) Aarti Singh (@CarnegieMellon) shows that you can *maybe* use a simple #NeuralNetwork to do that!
1
5
38
At Physics4AI workshop at Aspen this week, it is an awesome workshop, learnt so much!! . Site here: (talk slides will be up soon): 1/ .Theory of everything for Large scale deep learning! . mu-transfer @TheGregYang [how you transfer hyperparameters] . 1/n
1
1
39
Simply awesome (very biased π
) demonstration of how we can use machine learning for science! @PabloLemosP @MilesCranmer @Niall_Jeffrey @PeterWBattaglia.
Could machine learning rediscover the law of gravitation simply by observing our solar system?. With our new approach, the answer is *YES*. Led by: @PabloLemosP .With: @Niall_Jeffrey @cosmo_shirley @PeterWBattaglia.Paper: Blog:
1
2
36
It was really great to co-present with @MilesCranmer today at #NeurIPS21 on a tutorial on ML for Physics and Physics for ML! . Let's hope to see everyone in person at #NeurIPS22π€.
0
0
35
We are organizing the Mathematical and Scientific #MachineLearning Conference in @Princeton 2020! Submission deadline coming up! Nov 30: .
1
7
34
Very π€ͺ about our new paper in #ICLR2022! . Led by @neuro_kim, along with co-authors @DrumBushField,@dkochkov1, @MilesCranmer, @spectralhippo, Jonathan Godwin, Can Cui, @PeterWBattaglia and Alvaro Sanchez-Gonzalez! . Come see our poster!
We have a paper in ICLR! The title is βLearned Coarse Models for Efficient Turbulence Simulation.β We wanted to see if we could train general-purpose ML models to predict turbulent dynamics accurately at low spatial and temporal resolutions (1/n).
0
5
34
Come join us at @SimonsFdn and @FlatironCCA this coming year to work on foundation models for science! . with @MilesCranmer @kchonyc @oharub @Tim_Dettmers Michael Eickenberg, Siavash Golkar and more! . Applications are open now for positions at the faculty level and internships!.
Job alert! π¨. We are building a *Foundation Model for Science*. @SimonsFdn + @FlatironCCA are supporting PhD internships + faculty sabbaticals!. w/ @cosmo_shirley @kchonyc @Tim_Dettmers @oharub ++. Interested in building "ScienceGPT" with us? Please apply! (links in 2nd tweet)
1
14
30
Great lunch discussion at #neurips19 with @ylecun @PeterWBattaglia @KyleCranmer @MilesCranmer and more!.
0
2
34
A bit emotional that the telescope we worked on for 10?! years is finally taking off π». What will the Universe tell us? #EuclidMission . The Dark Universe Is Waiting. What Will the Euclid Telescope Reveal?
0
1
33
Ringed Uranus imaged by #JWST! . I never thought of Uranus as a ringed planet seriously until today! .
1
1
28
Potent points from our #NeurIPS2020 Machine Learning for Physical Sciences workshop: . Our panel speaker @iamstarnord asked."What is our role as scientists in developing algorithms for science, but which can also be used for things like the oppression and #genocide of #Uighurs?β.
Rather than a longer Twitter thread, here's an article in @Nature worth reading about βthe ethical questions that haunt facial-recognition researchβ which discusses some of these issues in depth, with several front line AI researchers with different views.
0
5
29
Interviewed by @TerryGlavin on the recent massive political movement in Hong Kong. I hope to give a voice to the students, to the minorities, to the underdogs in this story. #SaveHongKong #HongKongProtest #HongKongExtraditionLaw.
0
9
29
π₯ @PolymathicAI presents the coolest (IMO) and most diverse fluid dynamics dataset The Well at #NeurIPS2024 !. Date: tomorrow (Thursday) Dec 12 .Time: 11am-2pm (PST)!.Where: West Ballroom A-D #5102 . Led by @mikemccabe210 and @oharub !! . Data + code:
1
6
30
Calling all #MachineLearning peeps who are interested to do some #Astrophysics, finding the fundamental truths about our #Solarsystem , the #MilkyWay and our #universe! .We are hiring at @FlatironCCA, check out:
1
20
26
Proud of our new architecture that conserves energy in a learned simulator. Wonder what kind of #physics and #RealWorld problems we can solve with this new architecture? @MilesCranmer @samgreydanus @shoyer.@PeterWBattaglia @DavidSpergel.
1/10 Very excited to present Lagrangian Neural Networks, a new type of architecture that conserves energy in a learned simulator without requiring canonical coordinates. w/ @samgreydanus, @shoyer, @PeterWBattaglia, @DavidSpergel, @cosmo_shirley:
0
5
28
Causal graphs (merger trees) of galaxy formation are important for the prediction of galaxy properties! . Our recent paper led by @AstroCKragh showed how it is done with Graph networks π₯.
Graph Networks show huge potential for physics. But, in astrophysics, are there any *true* graph structures?. YES! Causal graphs ("merger trees") of galaxy formation!. with @MilesCranmer @peter_melchior @cosmo_shirley Rachel Somerville @a_gabrielpillai
0
2
27
We are excited that our paper is published at @NatureAstronomy! . More details will be posted soon on this series of papers by SimBIG collaboration π₯³. Stay Tuned!.
By extracting non-Gaussian cosmological information on galaxy clustering at non-linear scales, a framework for cosmic inference (SimBIG) provides precise constraints for testing cosmological models. @ChanghoonHahn @cosmo_shirley @DavidSpergel et al.:
0
4
26
You can now use Machine Learning to trace how stardust π« trace the underlying gas distribution of the Universe, accelerating these stardust simulations by many folds! . Great Work led by Timothy Yan-Mong Chan, @astroNatascha @philip_armitage :) !!.
New paper! In work led by Yan-Mong Chan, including @astroNatascha and @cosmo_shirley @FlatironCCA, we explore the use of deep learning to model how particles respond to fluid turbulence. (1/9) .
1
7
26
. @PolymathicAI's new work shows . How causal transformers outperform bidirectional transformers even for non-causal tasks! . Read more below π.
SOTA models often use bidirectional transformers for non-NLP tasks but did you know causal transformers can outperform them even on tasks without a causal structure?. Our recent work shows causal transformers learn circuits bidirectional ones can't, leading to better performance!
1
2
27
Cool diffusion model for science application! . Now these diffusion models are not only for hype :).
We present Folding Diffusion: a diffusion model for protein structure inspired by physical protein folding. Let by @Kevin_E_Wu during his internship, with @alexijielu @vdbergrianne @james_y_zou @avapamini . Code: Preprint:
1
4
25
On the @NASA Astrophysics Advisory committee today, and got reminded how π₯³cool NASA is ! . Working with our partners, NASA launched two telescopes this year @ESA_Euclid (from @NASAKennedy ), and @XRISM_jp (from Tanegeshima Japan)!!
1
1
25
What do you think foundation models should do for your science topic? @PolymathicAI @MilesCranmer @demishassabis @ylecun @SiavashGolkar @meickenberg.
3
1
23
Can't believe I will be missing #ICML in Hawaii π , but my teammates are there! . Check out talks/posters/workshops by π₯@changhoon_hahn @PabloLemosP Francois Lanusse, Liam Parker, @oharub, @meickenberg and @kchonyc . Guess where I am instead of Hawaii?
3
5
24
Excited that @MilesCranmer just passed generals at Princeton, moving into PhD candidacy. Looking forward to doing lots of cool science and ML and everything in between with @MilesCranmer, @PeterWBattaglia and @DavidSpergel !!.
1
0
23
Come join us at #NeurIPS2020 workshop "Interpretable Inductive Biases and Physically Structured Learning"! . Links to the workshop: Great poster talk by Kimberly Stachenfeld, Jonathan Godwin and @PeterWBattaglia on spectral #GraphNet
0
5
23
Doing cosmology and machine learning while feeling very "out of distribution" today. After >~2 years of not conferencing . this feels surreal. Note that >98% of people are not masked. #euclid2022 #oslo
0
0
24
I won't be able to do justice to what transpired at our symposium yesterday, but here is the TLDR: . Excited ! @neuro_kim told us about how one uses #ML to predict rigid body dynamics (which people thought was not possible before)! . And adding noise to the data actually help?!
1
1
24
Come hear my @SimonsFdn lecture if you are in town :).
CCA Cosmology X Data Science group leader Shirley Ho (@cosmo_shirley) will be giving a @SimonsFdn lecture titled "The First AI Simulation of the Universe is Fast and Accurate β and We Don't Know Why". 2/26 @ 5pm ET, NYC. @SF_Lectures #AI #Astrophysics
0
5
21
How do you weigh clusters of galaxies?. With Deep Learning + Symbolic regression! .New @PNASNews paper in Deep Learning X the Universe led by @JayWadekar1 ! . See the tweet below π.
In our new PNAS paper, we use machine learning to discover novel equations for deriving masses of clusters of galaxies from observational quantities! (1/10). with @MilesCranmer @paco_astro@jcolinhill @DavidSpergel @cosmo_shirley,Leander,Nick,Daniel,Lars.
0
4
23
Proud to be on the team for this work! @MilesCranmer @astrodantamayo @hannorein @PeterWBattaglia Samuel Hadden @philip_armitage and @DavidSpergel.
A group of astrophysicists including #PrincetonU's @MilesCranmer, @astrodantamayo, @cosmo_shirley and @DavidSpergel have used #MachineLearning to predict when a general planetary configuration will become unstable.
0
0
22
Kudos to @marinkazitnik for shepherding our @Nature paper of ML for science to completion!! π₯. I hope it is a short and interesting read for those interested what AI can do for science:.
Excited to share our @Nature paper on the role of AI in scientific discovery ππ¬ #AI4Science. AI is transforming discovery across sciences π€π From hypothesis generation to data interpretation, it is reshaping all stages of research in ways we could not imagine using
0
4
22
We are hiring! Want to work in the most vibrant (a bit biased opinion) city of north America on cutting-edge research?. Look no further! We @FlatironCCA are hiring postdoctoral fellows! . #Astrophysics #DeepLearning.#HiringNow .
0
6
21
Come hear @MilesCranmer present our paper "Learning Symbolic Physics with Graph Networks" at #NeurIPS2019 tomorrow! . Paper at:
Machine Learning and the Physical Sciences workshop at #NeurIPS2019 is tomorrow 14 December, West 119 & 110! Schedule: Live streaming:
0
7
22
Very excitedπ€ͺ to be giving the "Data for Good seminar" at @DataSciColumbia tomorrow!. "From Planets to the Universe: How Deep Learning Changes Science one small step at a time" . More info: .
1
3
21
Very wise statement. though I will still celebrate today since we are so done with being on overleaf non-stop for last X days due to #neurips2020.
As NeurIPS deadline nears, a wise person said:. When looking back at your career, you really don't want βK papers at NeurIPSβ to summarize your contribution to the field. You want your ideas to be so useful that most will just take them for granted and not even bother to cite you.
0
1
21
A little bright spot in this crazy year:. A little rover on mars called Perseverance! . #ι¦ζΈ―δΊΊζδ½ #HangInThere
0
0
20
Very proud to present our work at @PNASNews led by @astrodantamayo and @MilesCranmer along with @DavidSpergel and @philip_armitage on using #MachineLearning to predict which planetary systems will survive! . See paper at: .
A new @PNASNews paper shows how #MachineLearning methods can predict the stability of planetary configurations 100k times faster than previous approaches. Co-authors include CCA director @DavidSpergel and group leaders @cosmo_shirley and @philip_armitage.
1
6
20
If you are interested in ML or ML X Science postdoctoral fellowship positions. Join us at the @FlatironInst!! . @FlatironCCA fellowship (Oct 15 deadline) @FlatironCCM fellowship (Jan 1 deadline) : .
If you are on the postdoc market in ML X Science, come work with @cosmo_shirley's group at Flatiron Institute/Simons Foundation in NYC!. Deadline: October 15, 2020. Deadline: Jan 1, 2020.
0
13
19
Fantastic workshop for the Machine Learning X Physical sciences folks :). for the submission deadline, watch:.
I'm thrilled that our Machine Learning for Physical Sciences #NeurIPS2020 workshop proposal was accepted! Last year we had a great turnout and fantastic talks - check out the #ML4PS hashtag. New to team: @adjiboussodieng @iamstarnord @glouppe @zdeborova
0
5
20
Spending a sabbatical at @FlatironCCA is awesome, why? . You got to hang out with rocket scientists π, astrophysicists, AI scientists, blackhole simulators, and planet-hunters !! . and all within the vibrant city of NYC! π₯. Deadline to apply: Jan 2, 2024!.
Applications for sabbatical or long-term visits to CCA are now open for the Fall 2024 and Spring 2025 year. Learn more about the program:
0
2
16
Great to present our paper led by @LucasMakinen!. Note for the #MachineLearning peeps: the signal to noise level in these observations is incredibly low . .
STOKED to share our new paper βdeep21: a Deep Learning Method for 21cm Foreground Removalβ. We present an approach to separate 21cm cosmological maps from foregrounds -- no power spectra needed!. paper: tutorial: . a thread π.1/n
0
5
16
Excited by our recent work on Bayesian Neural network on simulation based inference led by @PabloLemosP and @MilesCranmer !. Come see our poster at #ICML2022 at the ML4Astro workshop!. Paper: Workshop:
Can Simulation-Based Inference analyses be improved by Bayesian Neural Networks?. In our recent paper in the ML4Astro #ICML2022 workshop, we show that they can, particularly in cases where the simulations fail to model every aspect of the data π§΅.
1
4
18
Happy to see Joanne Cohn, a cosmologist at Berkeley (ex-string theorist) gets the deserved recognition for starting @arxiv! .
0
2
18
Excited to see how we can now leverage large language model to accelerate the scientific process !! . We next need fully automated π€ building of physical experiments and data collection!.
The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery π©βπ¬. Itβs common for AI researchers to joke amongst themselves that βnow all we need to do is figure out how to make AI write the papers for us!β but I think weβre now getting there!.
2
2
20
I was pleasantly surprised by having the honor to be on the full female panel along with @radastrat and @spaceroboticist! .Even better, our topic rocks: .Deploying AI to understand the mysteries of the Universe. World Summit AI #WSAI22. Picture: thanks to @DawnHunter83
0
1
18
Super excited for #NeurIPS2020 workshop on Machine Learning for Physical Sciences!. Special kudos to @atilimgunes and @glouppe for the insane amount of work they have put in to make this happen!.
We've updated our the website for Machine Learning in the Physical Sciences with the schedule and more than 150 contributed papers! Many thanks to @glouppe and @atilimgunes for the website updates, all the organizers and all the reviewers! #ML4PS2020 .
0
4
17
Extremely pleased to see my fellow cosmologist in our group and academic sibling @jcolinhill named a #SloanFellow today!.
I am very grateful and excited to be named a 2022 #SloanFellow! I am deeply thankful to my mentors, collaborators, and the fantastic students and postdocs I am so fortunate to work with @ColumbiaPhysics and @FlatironCCA. Stay tuned for exciting cosmology to come!.
1
1
16
#AI4Space 1st Workshop on AI for Space.In conjunction with CVPR 2021! . Submission deadline March 6, 2021! .
0
6
17
Happy to announce that our work on creating Universe really really fast with AI which generalizes dramatically at @CCA with Siyu He and @BerkeleyPhysics had been mentioned as one of the 20 best work of the year by @physorg_com !.
Top @physorg_com articles include work from CCA's @cosmo_shirley & Siyu He, who used artificial intelligence techniques to generate complex 3-D simulations of the universe.
1
3
18
Got engineering chops to drive scientific discovery with Foundation Models? . The deadline to apply is today:
Have you all heard about ChatGPT or foundation models but want to build more than a chatbot with AI? . π₯ We at @PolymathicAI are building foundation models for science π₯ . Join us (@albertobietti @MilesCranmer @kchonyc @meickenberg @SiavashGolkar Francois Lanusse,
2
5
17
π€―the wave of demo roll-out from @OpenAI ! . I am most excited by the math-tutoring demo, how about you?.
Say hello to GPT-4o, our new flagship model which can reason across audio, vision, and text in real time: Text and image input rolling out today in API and ChatGPT with voice and video in the coming weeks.
4
2
17
Wondrous image (but from our real universe) from #JamesWebbSpaceTelescope! . After a year of impressive AI-generated images, it is surprising that a real image still manages to take our breaths away. Story:
0
1
17