Shuiwang Ji
@ShuiwangJi
Followers
2,561
Following
3,136
Media
3
Statuses
1,636
Machine Learning, AI for Science, Professor and Presidential Impact Fellow, Texas A&M University Fellow, IEEE and AIMBE
College Station, TX
Joined June 2012
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
オーロラ
• 512991 Tweets
#Eurovision2024
• 378867 Tweets
Joost
• 346630 Tweets
#الهلال_الحزم
• 128912 Tweets
Forest
• 124323 Tweets
الدوري الاقوي
• 83360 Tweets
Reece James
• 59804 Tweets
الدوري السعودي
• 55519 Tweets
Ali Koç
• 51286 Tweets
Brahim
• 43372 Tweets
الدوري التاريخي
• 42659 Tweets
Granada
• 41228 Tweets
Arda Güler
• 39335 Tweets
Burnley
• 37697 Tweets
Rohit Sharma
• 34291 Tweets
#الدوري_الاقوي_في_التاريخ
• 26724 Tweets
DiniKonferansa TehditVeEngel
• 26400 Tweets
Luton
• 25874 Tweets
Sterling
• 20661 Tweets
Mudryk
• 18653 Tweets
#KKRvsMI
• 16917 Tweets
Poch
• 14094 Tweets
Caicedo
• 14064 Tweets
نيمار
• 14025 Tweets
Odoi
• 11255 Tweets
كوليبالي
• 10409 Tweets
New paper alert:
Self-Supervised Learning of Graph Neural Networks: A Unified Review
A comprehensive code library is coming in a few weeks. Stay tuned...
2
77
335
Excited to share our latest survey paper: "Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems"! 🚀 ArXiv: .
5
73
255
Two ICML papers on GNN explainability and generation got accepted, with code available:
On Explainability of Graph Neural Networks via Subgraph Explorations
GraphDF: A Discrete Flow Model for Molecular Graph Generation
1/2
2
9
137
26-page unified review paper on SSL of GNNs, accepted to TPAMI:
Open-source code
3
25
92
NeurIPS 2021 Track Datasets and Benchmarks Round2 AC recommendation. Very hard to see any logic:
"While all reviewers agree on acceptance, this is still a borderline paper, tending to reject, due to the lack of enthusiasm from the reviewers."
7
11
92
interested in graph OOD problems?
GOOD: A Graph Out-of-Distribution Benchmark
1
17
75
new ICML paper GraphFM: Improving Large-Scale GNN Training via Feature Momentum
0
13
77
DIG: A research-oriented library that includes unified and extensible implementations of algorithms for (1) graph generation, (2) self-supervised learning on graphs, (3) explainability of graph neural networks, and (4) deep learning on 3D graphs.
1
23
74
🔬 An in-depth yet intuitive discussion on symmetry, as well as explainability, out-of-distribution generalization, large language models, and uncertainty.
📖 Access categorized lists of resources to enhance learning and education.
🌐 Website: .
0
17
70
Interested in GNN and molecular property prediction? We have been working on a major project with many talented students. The result is a comprehensive ML/DL software package for graphs, sequences, and molecular property prediction, combining new methods with turnkey software.
2
11
69
I will be at NeurIPS 12/12-12/16 and speak at the Generative AI workshop. Excited to chat with anyone on AI/LLMs for science.
1
4
72
New paper on SSL for Graphs:
Group Contrastive Self-Supervised Learning on Graphs
Paper link:
1
13
58
Proud to join other distinguished colleagues as a Distinguished Member of ACM.
2
0
42
Delighted and honored to be elected to the AIMBE College of Fellows.
Congratulations to the 153 individuals elected to the AIMBE College of Fellows Class of 2022! We're so glad to welcome you to the AIMBE community. You can browse the full list of the new cohort, here:
5
17
120
3
2
39
New work on Crystal Material Property Prediction using transformer, accepted to NeurIPS 2022.
Periodic Graph Transformers for Crystal Material Property Prediction
by
@KeqiangY
et al. including
@lost_his_way
,
@ShuiwangJi
#ComputerScience
#Learning
0
6
20
3
5
32
Interested in self-supervised learning and denoising? A demo of our NeurIPS paper Noise2Same
by Abubakar Abid
@abidlabs
. Thank you!
1
7
26
Interested in building deeper GNN? here you go.
Glad to share our
#kdd2020
work "Towards Deeper Graph Neural Networks". Joint work with
@xiao9aba2008
and
@ShuiwangJi
.
Paper:
Code:
(0/4)
1
7
32
2
8
23
The Department of Computer Science and Engineering at Texas A&M University invites applications for two full-time tenure-track or tenured positions in the area of Data Science.
0
10
24
Will be at NeurIPS next week and giving an invited talk to the AI for Science Workshop () on Modeling Complete Potentials in Crystal Materials
0
1
24
Check out our work accepted to NeurIPS 2020, including paper, code.
Noise2Same: Optimizing A Self-Supervised Bound for Image Denoising
pdf:
abs:
github:
0
5
46
0
1
23
highlights:
🌟 A 263-page paper by 63 authors from 14 institutions, including 41 figures and 36 tables.
🔍 A spectrum of scales: subatomic (wavefunctions, electron density), atomic (molecules, proteins, materials, interactions), and macro (fluid, climates, subsurface).
0
2
23
Code is released as part of the DIG: Dive into Graphs
Any feedback is greatly appreciated.
2/2
0
6
17
LoG 2022 Reviewer Sign-up Form
0
3
18
Line Graph Neural Networks for Link Prediction
Line Graph Neural Networks for Link Prediction
by Lei Cai et al. including
@ShuiwangJi
#DeepLearning
#ComputerScience
0
3
5
0
3
14
Our new work on Augmented Equivariant Attention Networks for Electron Microscopy Image Super-Resolution
Augmented Equivariant Attention Networks for Electron Microscopy Image Super-Resolution
by Yaochen Xie et al. including
@ShuiwangJi
#AttentionModels
#DeepLearning
0
0
1
0
4
14
New paper alert: Non-Local Graph Neural Networks
0
2
13
Our new work accepted to KDD 2020 on dramatically accelerating attention models on high order data.
0
6
25
1
1
12
Capturing long-range info in GNN? check our recent work
The reason is the exponentially-growing receptive field, that squashes exponentially-growing information into fixed-length vectors.
If you want to pass a message to a distance of K - the cost is O(degree^K) messages that are squashed into a single vector.
(2/n)
1
0
6
0
4
12
We have a complete explanation of PCA/SVD using only linear algebra. If you are teaching PCA/SVD, try our material
This is amazing. Have not seen this type of true excitement for SVD since my PhD days with my advisor :-) cc
@vagelispapalex
@PoloChau
@danaikoutra
@alexbeutel
1
0
8
1
1
11
@NeurIPSConf
@kchonyc
Will reviewer-AC discussions be visible to authors? If yes, will authors be able to respond to discussions if necessary? SO many authors have worked very hard on rebuttal, but many reviewers did not even click the ACK box.
2
2
12
@jure
@BaiduResearch
@DeepMind
@Synerise
Thank you for organizing this great event. We are excited to be one of the few teams from academia on the winner's list. Our tool is available at which also include code and tools to achieve
#1
on the AI Cures
0
1
12
New paper alert: Data Augmentations for Graphs
Automated Data Augmentations for Graph Classification
by
@LuoYouzhi
et al. including
@ShuiwangJi
#Statistics
#Probability
0
3
15
0
1
12
Interested in Interpretable Metric learning? Here you go: Towards Improved and Interpretable Deep Metric Learning via Attentive Grouping
Towards Improved and Interpretable Deep Metric Learning via Attentive Grouping
by Xinyi Xu et al. including
@ShuiwangJi
#EvaluationMetrics
#ConvolutionalNeuralNetwork
0
0
3
0
0
11
Will be at NeurIPS next week and present
Periodic Graph Transformers for Crystal Material Property Prediction
0
0
12
Learning From Data book, Amazon‘s bestseller in
#MachineLearning
, builds a solid foundation concisely and crisply.
0
1
11
Our new work on DL for image transformations.
Global Voxel Transformer Networks for Augmented Microscopy
by Zhengyang Wang et al.
#DeepLearning
#NeuralNetwork
0
24
23
0
1
11
@SergeyI49013776
Ultimately research should be evaluated by how it impacts science and society, not where it was published.
0
0
11
NeurIPS: I have a reviewer who raised a few concerns a few days ago. Most of these concerns are caused by misunderstandings. Once we pointed these out in response, this reviewer lowered his/her score without even responding with a word. What can we do?
2
0
10
2/2
MoleculeKit: Machine Learning Methods for Molecular Property Prediction and Drug Discovery
Paper:
Code:
0
3
10
0
1
9
Interested in GNN and pooling on graphs? Built on our work on graph U-Nets, we push the field one step forward by incorporating graph topology. Enjoy.
Topology-Aware Graph Pooling Networks
by Hongyang Gao,
@lost_his_way
, and
@ShuiwangJi
#NaturalLanguageProcessing
#ComputerVision
0
0
2
0
2
7
Unspoken, but unmistakable, is one more question: Am I welcome?
At MIT, the answer, unequivocally, is yes.
0
1
9
The Department of Computer Science and Engineering at Texas A&M University invites applications for multiple full-time tenured or tenure-track positions. Open rank, open area.
0
6
9
@chaitjo
I think the major challenge of apply GDL to physics and science problems are how to encode physics priors like symmetries into the model in an flexible and efficient manner. Of course the GNN community might have a different focus/challenge.
1
0
9
Congrats Yi Liu for the great job!
We have new leaders in the synaptic cleft detection category of the
#CREMIChallenge
. Congratulations to groups DIVE and CleftKing!
1
1
3
0
0
7
@wellingmax
@jbrandi6
@robdhess
@ElisevanderPol
@erikjbekkers
Great to see a leap after our submission, which is
#2
now.
0
1
8
@ylecun
Small batch size makes learning quick, but could be noisy. Quick science might be good, but is too noisy. If I have to make a binary decision between quick and slow science, I will chose slow.
2
1
8
My 3DCNN work has been ranked as one of TPAMI's top 5 most popular articles since June 2016.
0
0
7
Join us for the KDD Cup OGB-LSC Workshop
0
0
7
New paper alert: iCapsNets: Towards Interpretable Capsule Networks for Text Classification
0
1
6
Second-Order Pooling for Graph Neural Networks, TPAMI
Second-Order Pooling for Graph Neural Networks
by Zhengyang Wang et al.
#NeuralNetwork
#ComputerScience
0
14
41
0
3
6
@LogConference
Please try to be inclusive and accept all good papers. This area is burgeoning with a lot of good papers each year.
1
0
7
1
0
5
@beenwrekt
One cause is the large number of submissions, leading to large numbers of (thus potentially noisy) reviewers, AC etc. How about limit the number of submissions under each name?
1
0
7
Looking forward to the ICML 2023 machine learning conference starting July 17.
0
0
6
Action and decisions are reversed, but hurt is done and can never be reversed.
DOJ just ended the China Initiative for "broader approach" to counter threats incl foreign exploitation of US science. I write
@WIRED
why fixation on borders & national interest obscures fundamental ethical concerns & incurs much more profound loss on all:
3
32
137
0
1
6
Our new ICLR paper:
StructPool: Structured Graph Pooling via Conditional Random Fields
0
2
6
@SimonShaoleiDu
We have a lecture notes that formally describes the relations between DL and kernel methods.
0
0
6
@GalChechik
Both the author (failed to disclose COI) and the reviewer should be banned from this and related conferences for N years.
0
0
6
@ylecun
@ericxing
@YiMaTweets
There is randomness in any review process, and limiting the number of accepted papers may only increase randomness. I found many of the criticisms we received during peer review are constructive, and addressing them led to better quality work.
1
0
6
Nice. Although I am not aware of a formal paper/document on this, I found there have been some discussions and hints on the relations. I compiled a lecture note used in my DL/ML class here
Astonishing - deep networks are just kernel machines, regardless of architecture:
14
54
264
0
2
5
@mmbronstein
There seems to be quite significant mismatch between papers and reviewers this year. We got a generic, one-sentence comment for a paper.
1
0
5
Will be at NeurIPS next week and present
ComENet: Towards Complete and Efficient Message Passing for 3D Molecular Graphs
0
1
5
@chaitjo
If there is a known prior, you better encode it explicitly, rather than letting data and optimization figure out. If optimization works perfectly, do we need ResNet?
0
0
7
@tdietterich
@percyliang
“Foundation Models” is certainly not my favorite name as it carries zero info.
1
0
4
@kchonyc
@NeurIPSConf
Any reason for excluding authors? The author-reviewer discussions are not mediated by AC/SAC and thus many reviewers did not even read rebuttals. It will be beneficial to allow author participation during reviewer-AC discussions.
0
0
5
@peter_richtarik
Very true and the root cause is the large number of submissions. Either group leaders should do their QC or conferences should put a limit on the number of submissions each author can do.
1
1
5
0
0
4
@AnimaAnandkumar
@jo_brandstetter
@rejuvyesh
@Mniepert
Not a surprise. We have worked very extensively on all three and have tried to optimize everything we can. One thing we observe was that conclusions can be overturned easily if models are not optimized properly.
0
0
4