Shuiwang Ji
@ShuiwangJi
Followers
3K
Following
6K
Media
14
Statuses
2K
Machine Learning, AI for Science, Professor and Truchard Family Chair, Texas A&M University
College Station, TX
Joined June 2012
RT @ShuruiGui: Any interests in extracting invariant functions or physical laws from complex environments? Check our new ICML paper โDiscovโฆ.
arxiv.org
We consider learning underlying laws of dynamical systems governed by ordinary differential equations (ODE). A key challenge is how to discover intrinsic dynamics across multiple environments...
0
1
0
5. More details can be found in the preprints:.PubChemQCR: TDN: MLIP Foundation Model:
arxiv.org
Accurate molecular property predictions require 3D geometries, which are typically obtained using expensive methods such as density functional theory (DFT). Here, we attempt to obtain molecular...
0
1
5
4. MLIP Foundation Model: Existing MLIP models pre-trained on PubChemQCR can serve as foundation models that can either generate low-energy geometries or be fine-tuned for downstream tasks to boost molecular property prediction.
0
0
0
3. TDN: A fast, approximately SO(3)-equivariant MLIP model using tensor decompositions, with massive speedups โก and SOTA performance on PubChemQCR, OC20, and OC22.
0
0
0
2. PubChemQCR: We introduce the largest DFT-based relaxation dataset, comprising 3.5M unique molecular relaxation trajectories and 300M+ conformers with energy & force labels, in which 105M+ conformers are DFT-calculated.
0
0
0
Excited to share our recent work on building scalable, efficient, and accurate machine learning interatomic potentials (MLIPs) for molecular simulation!
4
2
24
New paper alert: Mitigating Spurious Correlations in LLMs via Causality-Aware Post-Training.
0
2
17
New paper alert: Efficient Prediction of SO(3)-Equivariant Hamiltonian Matrices via SO(2) Local Frames.
0
0
13
New paper alert: A Two-Phase Deep Learning Framework for Adaptive Time-Stepping in High-Speed Flow Modeling.
0
0
4
New paper alert: Curriculum Reinforcement Learning from Easy to Hard Tasks Improves LLM Reasoning.
1
3
23
RT @KeqiangY: Excited to share our latest preprint on LLM Agents, beyond tool use, for Materials Discovery. ๐. โจUnleash the discovery powerโฆ.
0
5
0
New paper alert: Toward Greater Autonomy in Materials Discovery Agents: Unifying Planning, Physics, and Scientists.
1
6
32
RT @tdietterich: @tallinzen @F_Vaggi Teaches American kids the skills they need to cure diseases or start companies. She embodies the Ameriโฆ.
0
1
0
Science is under siege. Trumpโs latest Executive Order calls for politically appointed science commissars to evaluate research. Join me in adding your name to our open letter condemning Trumpโs escalated assault on science in America. Sign here:
actionnetwork.org
Science is under siege. Join us in adding your name to our open letter condemning Trumpโs escalated assault on science in America.
0
0
5
RT @_vztu: ๐ ๐๐ฒ๐ป๐ฒ๐ฟ๐ฎ๐๐ถ๐๐ฒ ๐๐ ๐ถ๐ ๐๐ต๐ฒ ๐ก๐ฒ๐
๐ ๐ค๐๐ฎ๐ป๐๐๐บ ๐๐ฒ๐ฎ๐ฝ ๐ณ๐ผ๐ฟ ๐๐๐๐ผ๐ป๐ผ๐บ๐ผ๐๐ ๐๐ฟ๐ถ๐๐ถ๐ป๐ดโ๐๐ฒ๐ฟ๐ฒโ๐ ๐๐ต๐ฒ ๐๐ถ๐๐ฐ๐ต๐ต๐ถ๐ธ๐ฒ๐ฟ'๐ ๐๐๐ถ๐ฑ๐ฒ ๐ฌ๐ผ๐โ๐๐ฒ ๐๐ฒ๐ฒ๐ป ๐ช๐ฎ๐ถ๐๐ถ๐ป๐ด ๐๐ผ๐ฟ! ๐. We'reโฆ.
0
47
0
Texas A&M Hosts Inaugural AI Workshop For Science And Engineering.
0
0
11
RT @GalantiTomer: I really enjoyed the RAISE Workshop ( at @TAMU today and share my work! Huge thanks to Prof. @Shuโฆ.
0
1
0
We have 5 papers accepted to ICML. I am especially happy for Monte, an undergraduate student with the second paper as the primary author (the first one was ICLR spotlight).@icmlconf @MBohde21912.
4
1
78