Sergei Kalinin
@Sergei_Imaging
Followers
3,698
Following
1,416
Media
302
Statuses
6,137
Weston Fulton Chair professor, UT Knoxville: using the power of (the Force) ML for moving atoms and learning physics ex Amazon, ex ORNL
Joined January 2015
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
オーロラ
• 508456 Tweets
Joost
• 344098 Tweets
Forest
• 116459 Tweets
#الهلال_الحزم
• 113550 Tweets
الدوري الاقوي
• 63224 Tweets
Modric
• 59387 Tweets
Ali Koç
• 49697 Tweets
Dremo
• 44621 Tweets
الدوري السعودي
• 43442 Tweets
Burnley
• 36022 Tweets
Granada
• 35004 Tweets
الدوري التاريخي
• 31098 Tweets
Brahim
• 30419 Tweets
Arda Güler
• 27114 Tweets
Luton
• 24332 Tweets
#NfoChe
• 22037 Tweets
Aina
• 21125 Tweets
#الدوري_الاقوي_في_التاريخ
• 18018 Tweets
DiniKonferansa TehditVeEngel
• 17616 Tweets
Reece James
• 15798 Tweets
Mudryk
• 15353 Tweets
Sterling
• 14783 Tweets
نيمار
• 11605 Tweets
And at Chapter 7 it starts to become truly useful and interesting. But it's time to switch to regular daily activities...
4
72
632
The new language model for science . Upon few quick tries, it seems to generate professional text in the areas I am familiar with. And 7 years ago we were *joking* about ML writing papers!
9
57
365
I am standing amazed of my long-term collaborators in Kiev that just submitted the revised version of our joint paper....
2
11
281
Feeling proud - my student told me that he is using ChatGPT to improve the paper. Will be first publication with the AI coauthor!
16
9
202
SVK: h-index (Google Scholar) -> 100. I wonder if this is the first order phase transition:)?
10
1
160
@Sergei_Imaging
+
@Robo_Perovskite
+= Celine Ahmadi-Kalinina. New human in the loop ML project launched!
24
0
125
SVK -> Feynman prize for experimental work on ML and STEM atomic fabrication.
HUGE CONGRATULATIONS to our 2022 Foresight Institute Feynman Prize Winners! Sergei V. Kalinin
@Sergei_Imaging
for Experimental Work, James R. Chelikowsky for Theory, and Dr. Emanuele Pennocchio
@EPenocchio
Distinguished Student Award 🎉
2
2
34
23
4
93
Physics is the new data. Some analysis of trends in ML my colleagues
@MaximZiatdinov
, Bobby Sumpter, and
@andrewwhite01
made over last several years. Comments are welcome
@GaryMarcus
,
@ylecun
,
@yudapearl
4
15
85
@tunguz
As proposed by
@MaximZiatdinov
, there is no contradiction between code and paper. They can be combined in a single format, with text as markdown and code as code. Plus all the refs, etc. Then this document can be cited as paper, and code can be shared
1
4
71
Rama Vasudevan
@ramav_matsci
-> APS Fellow, "For pioneering and visionary development of open-sourced physics-based machine learning methods in atomic-scale and mesoscopic imaging, and their application in physics". Congratulations, Rama!
8
2
66
DFT gives some (often good) approximation of reality. ML on DFT allows to interpolate, but it's approximation of approximation. The challenge is for ML to update ML/DFT models based on experimental data.
3
8
63
Dear colleagues - if (a student) or (have a student) you would like to spend a year at ORNL to work on science driven machine learning based automated experiment in electron or scanning probe microscopy - let me know (check out Office of Science Graduate Student Research (SCGSR).
2
24
61
@ResNeXtGuesser
I reckon no normally trained DCNN classifier had example for banana water. Think about it, I am with the classifier on this issue.
1
0
56
This special day in life of scientist-parent when you son publishes first paper:)
8
1
57
New experience in scientific life - have paper rejected from a journal ( transfer offer), submit it to a different one and get very positive reviews, and get (out of the blue) acceptance (!) from the first journal.
6
1
54
Celine watches reinforcement learning tutorial in summer school for machine learning in electron microscopy given by Rama Vasudevan
2
1
53
First time I get review that says "This is a well-conceived, timely, and visionary article whose publication will be a central part of the revolution of elevating automation from a local optimizer to a global accelerator of scientific discovery”
3
2
50
SVK citations -> 45k. Always interesting to explore areas between disciplines - nanoscale electromechanics, machine learning in imaging and materials design, and direct atomic fabrication. And for all three of them the next decade looks like becoming mainstream!
2
2
50
As a professional recognition - I am honored to be invited to give the Manuel Cardona lecture of the The Institut Català de Nanociència i Nanotecnologia, ICN2.
5
4
51
Note for when using ML in physics. First, solve the problem using functional model, and then see if ML is even necessary....
1
2
48
First in-person microscopy conference. When meeting people, almost like 3 years didn't pass. But the science is totally different now!
2
0
46
SVK Career update (belated): ORNL -> Amazon (principal scientist, sabbatical)/University of Tennessee Knoxville (professor). My new email is sergei2
@utk
.edu
0
1
46
Looks like there starting to be a very strong resonance to recent Nature papers by Google Brain team and A-Lab.
One immediate thought:
- we as a community really invested a lot in theory
- and not really enough in inverse problems - analysis of scattering and spectral data
1
2
45
For ML, the possible next step beyond status data sets and SOTA chasing may be deployment on (almost) closed experimental systems exploring or optimizing materials with relatively simple physics including automated microscopes or automated synthesis, as
1
11
46
I wish to people who created EndNote for Word to use it themselves for big multi-author documents, and enjoy every moment of it.
4
1
43
Atomic-scale e-beam sculptor patent granted!
Took a while after it was filed - covers atomic motion, 3D crystallization, and feedback based on learned cause and effect relationships between beam and atomic changes.
2
1
42
The most remarkable thing about science is that it cannot be owned. If someone tries to control it in one place, it starts to develop in another one.
0
5
42
@MaximZiatdinov
Why? If followed, this principle will really open doors to very unfortunate outcomes (suppress collaboration, dirty politics, you name it)
1
0
43
We argue that the unique opportunity given by electron and scanning probe microscopies is to learn generative physical models (much like astronomers create models of universe based on observations).
@ramav_matsci
@MaximZiatdinov
1
4
40
Enter the hypothesis learning - aka active learning co-navigating theory and experiment via hybrid of reinforcement learning and structured Gaussian Processes
@MaximZiatdinov
0
2
40
And this is the condition they Steve Pennycook, the best mentor I had when at ORNL, made me sign as part of a deal to collaborate on STEM on functional materials. Hopefully 18 years later, ChatGPT will do it for me....
2
0
41
@J_Lovering
@lexfridman
You really assume that people choose books to read based on the gender (or nationality, or race) of the author?
2
1
36
Fantastic to see Ph.D. advisor as a paragon of women in science to admire. +1 on that!
0
1
39
Science is a gift economy. Citing is saying thank you.
1
2
37
Q: "Can machine learning discover new physics?"
A: "If we can understand it, it's not new. And if we cannot, it's not physics"
6
1
38
For the ML aficionados attending (real or virtual) 2021 MRS Fall Meeting -
@MaximZiatdinov
and yours truly will be giving tutorial CH04: Machine Learning and AI Methods for Materials Science—Applications to Imaging and Smart Experiments, 8:30 am – 5:00 pm ET, Monday, Nov. 29th.
3
3
38
My favorite reviewers comment is "This paper should be proofread by a native English speaker"
5
1
38
The realization of a fully autonomous system for physics discovery - for ferroelectric domain growth. Probably first practical realization!
@Yongtao_Liu
@MaximZiatdinov
@ramav_matsci
1
4
37
A weird feelings for a scientist - reading a paper in a high profile journal that essentially repeats (idea, motivation, material) your decade old paper in another high profile journal. For a very small community. Any voice of advice,
#AcademicChatter
and
#AcademicTwitter
?
6
1
36
Bridging high resolution electron microscopy and supercomputing as integrated workflow. Kudos to
@MaximZiatdinov
of AtomAI, and non- twitterized Ayana Ghosh, as well as Bobby Sumpter and Ondrej Dyck.
0
6
35
Reinforcement Learning for atomic manipulation in electron microscope - by
@KevinRoccaprio1
and Google Brain Team
@pcastr
@AaronCourville
@max_a_schwarzer
@ekindogus
and many others.
It was a decade late ng quest for me since our Atom Forge paper!
0
5
36
Bayesian methods for active experiment - from simple BO and inference to deep kernel learning and structural GP and hypothesis learning. Can apply to other spaces - including chemical or processing.
@MaximZiatdinov
@ramav_matsci
@Yongtao_Liu
1
5
36
Two weeks as UTK faculty - check
Two NSF proposals submitted - check
Two travel (foreign and domestic) - check
Time to get ID and parking pass!
1
1
35
@GaryMarcus
@GaryMarcus
, I think you are now crossing the boundary. Disagreement with ideas is one that ng. Declaring person espousing different views as "the most dangerous person" is a totally different thing. Shame on you
1
0
32
Pleasant surprise of the morning - responding to professional and in depth reviews for two papers, where reviews clearly understood the concept and offered comments that will make paper clearer. Whomever you are, thank you!
1
0
35
This special feeling when your high school senior son tells you that he has written for fun LSTM predictor for SMILES -> solubility using AqSolDB data set
0
2
34
Machine learning for better hybrid perovskites -> Joule. Great collaboration and brainstorm with
@Robo_Perovskite
,
@MaximZiatdinov
, Eric Lass , and Yuanyuan Zhou.
1
4
33
Publication is sharing ideas. By now, we understand that sharing data and code is important. I wonder if publications should also have an optional share of materials (the samples may be available upon request) and training. Science is built as a community.
3
2
33
Lectures from ORNL workshop on Automated Experiment and Machine Learning in Scanning Probe Microscopy -> online. Featuring lectures by Mikhail Katsnelson , Leroy Cronin, Danilo J. Rezende , and tutorials on VAEs, DKL, DCNNs, and PyCroscopy. See:
3
6
34
Nature Materials: ferroelectricity in hafnia controlled by the surface electrochemical states
0
5
34
Microscopy is all you need. To bring ML to physical world applications, scanning probe and electron microscopy can be a perfect playground - before automated cars and robotics.
@MaximZiatdinov
@ramav_matsci
@KevinRoccaprio1
@Yongtao_Liu
3
6
33
@FunkEntropy
@timgill924
Many stand in awe at accomplishments of Tim Gill
@timgill924
, the world first quantum sociologist. But many still have to arrive to enlightenment.
1
1
32
Meeting Andromeda, Kraken, and famous
@TEMgrazul
(courtesy of brief tour by
@David_A_Muller
). What else do they do at Cornell?
1
1
33
SVK -> Fellow, Asia-Pacific Artificial Intelligence Association.
5
1
32
This special feeling when:
- your postdoc gets a faculty offer in excellent school
- with a start-up twice of yours...
1
3
32
1. Baby looks particularly contemplative after proving that the mass conservation laws in fact apply to babies.
2. Sproggy is confused - socks this small are difficult to chew
0
0
33
An opinion on deep learning in electron and scanning probe microscopy - with
@ColinOphus
,
@eric_stach
,
@MostlyPhysics
,
@joshagar
,
@MaximZiatdinov
, Steven Surgeon, and legendary John Randall
1
3
28
@BenBlaiszik
@andrewwhite01
I think for all of us part of the problem that our research is well outside classical domain sciences, yet not fully in the ML field (with NeurIPS, ICML, etc). This community is still very latent
1
0
30
The most amazing thing about all things automated experiment. One needs to know at some level:
- domain area (to even start)
- ML (by now beyond scikit-learn)
- elements of workflow building (Colab)
- IoT Integration
- and ideally cloud technology (AWS)
Sounds like a challenge!
2
0
30
Automated Scanning Probe Microscopy
@Yongtao_Liu
and hybrid perovskites
@Robo_Perovskite
can go hand in hand. Stand by for AE SPM that explores conductivity and current-voltage hysteresis (aka ionic motion) on grain boundaries only. There is some clear internal variability!
2
0
28
The introduction into Gaussian Processes from two physicists perspective -> Medium.
0
1
29
Life is different in Seattle... I have never seen 30 scientists with throwing axes before...
2
0
28
During my thesis defense exactly 20 years ago, half of my thesis committee (from materials science) argued that my thesis should be physics, and the second part (from physics) argued that it is Mat. Sci. Go figure - it is still not clear if I do physics/MSE or machine learning.
1
2
28
Our first publication on machine learning in STEM/EELS in 2010. We have been at it for a while - and it's great to see how well these predictions materialize (albeit it took deep kernel learning to do structure -EELS matching well).
0
2
29
Hello, COVID. This particular date was delayed for 2.5 years, but now we meet. As the wise man said, این نیز بگذرد
10
0
29
And if you are interested in machine learning in material science - synthesis and microscopy - follow
@Sergei_Imaging
@MaximZiatdinov
@ramav_matsci
@Robo_Perovskite
and
@Yongtao_Liu
:) Also,
@timgill924
for voice of academic wisdom and hint of quantum sociology.
2
5
27
They call it Windy City for a reason. On a positive side, I got to meet multiple denizens of the academic Twitterverse in AI for materials in person!
@taylordsparks
1
0
28
This is a fundamental problem with the ML community now - trying to learn domain sciences directly, rather than building a collaborative network with domain groups that have been invested in (simpler) ML for a while and have translational expertise
I love when ML researchers get excited about science, but seriously the reviewing process for scientific applications at ML conferences (e.g., ICLR) is entirely broken.
Papers with glaring errors are sailing through, without a single review from somebody with domain expertise.
20
26
352
0
4
28
Sometimes the right physics discovery workflow is the simple physics discovery workflow. Congratulations, Mani!
1
2
28
Just got a review asking to justify that VAE is unsupervised method....
5
1
28
One thing for sure - with ChatGPT, my students who had minimal or zero programming experience, now do homework problems in my course that 3 years ago would be a paper in high profile journal (in fact, that's where the problem has come me from).
@sama
and team did change the world
2
4
27
DFT functionals should be probabilistic, and we should be able to propagate uncertainty to the calculation results.
2
1
25
@skwinnicki
At least she is not interested in electromechanical responses of cardiac miocytes
1
0
27
The first paper of the year with
@Robo_Perovskite
lab and building on the
@MaximZiatdinov
VAE libraries! For al ye hybrid perovskites and automated synthesis afficionados!
0
2
27
Speaking of review process - I wonder if it will be efficient to go from written reviews to scheduling an online anonymous Zoom session? There is no problem now to hide/alter voice. Can save time and make process fast and dynamic. If interesting - reshare!
#AcademicTwitter
5
0
27
It's important to make a post that communicates joy of life and does not offend sensibilities of Twitter owner.
2
0
27
For the ferroelectrics and PFM enthusiasts - now we have a capability to perform PFM spectroscopy when the temperature of the tip is dynamically changed. Read about it and contact Kyle Kelley to run on own samples via CNMS user program.
2
3
26
It is really time to stop considering pure "data driven ML" as a new or best way forward. Physics is the new data!
3
0
26
@fchollet
That's the first rule of using machine learning in experimental workflows: first do it without machine learning, and see what the bottlenecks are. May be they not even require machine learning!
0
0
26
The atomic construction community:
- half Russian-half Jewish
- Israeli
- Lebanese-Ukranian (Russian -speaking)
- Kurdish Iranian
We have the same goal!
1
0
26
Using Bayesian Optimization in the latent space of autoencoder trained on domain-specific examples. Here, we use it to create curly ferroelectrics - in theory so far. Will also work for materials and process optimization.
@MaximZiatdinov
@ramav_matsci
2
3
26
Meeting Javier Garcia Abajo - and learning what else can EELS do
0
0
26
And now ELIT DCNN for STEM. Converting data stream into atomic positions in real time as a plug-in for NION SWIFT. We can find them, we can explore how they respond to beam.
@KevinRoccaprio1
@MaximZiatdinov
2
2
26
Schools need courses on the history of science. And high schools on the history of science in the past 20 years.
5
2
24
The news from Iran are horrendous. Amazing not to see it front and center.
2
2
25