Conference Logo ProgramMaster Logo
Conference Tools for 2025 TMS Annual Meeting & Exhibition
Login
Register as a New User
Help
Submit An Abstract
Propose A Symposium
Presenter/Author Tools
Organizer/Editor Tools

About this Abstract
Meeting 2025 TMS Annual Meeting & Exhibition
Symposium AI/Data Informatics: Computational Model Development, Verification, Validation, and Uncertainty Quantification
Presentation Title A Complete AI-Accelerated Workflow for Superconductor Discovery
Author(s) Jason Gibson, Ajinkya Hire, Benjamin Geisler, Phil Dee, Peter Hirschfeld, Richard Hennig
On-Site Speaker (Planned) Jason Gibson
Abstract Scope In recent years, there has been a surge of interest in electron-phonon superconductors. However, a significant barrier in high-throughput screening of potential superconductors lies in the prohibitively expensive calculation of the Eliashberg spectral function (a2F), a key determinant of the superconducting critical temperature. Our past work on developing a bootstrapped ensemble of tempered equivariant graph neural networks (BETE-NET) has shown the ability of machine learning to bypass the costly a2F calculations. In this talk, I will detail the improved second-generation BETE-NET models, which were trained on an order of magnitude more data and obtained prediction MAEs of just 0.9K for critical temperature. I will then describe our AI-accelerated workflow that leverages elemental substitution, machine-learned interatomic potentials, and BETE-NET to identify novel superconductors. The preliminary search has already confirmed 30 candidates with critical temperatures greater than 5K, and among them, 8 with a critical temperature greater than 15K.
Proceedings Inclusion? Planned:
Keywords Machine Learning, Computational Materials Science & Engineering,

OTHER PAPERS PLANNED FOR THIS SYMPOSIUM

A Complete AI-Accelerated Workflow for Superconductor Discovery
A Generalizable, Accelerated, and Interpretable Artificial Intelligence Framework for Predicting Evolution of Materials Microstructure
A Generative Deep Learning Initialization Strategy to Accelerate Genetic Algorithms for Crystal Structure Prediction
A Machine Learning Approach to Optimize T5 Heat Treatment Conditions for Al-Si Alloys
A Materials Informatics Approach to Quantify Processing – Structure – Property Relationship in Boron Carbide Ceramics
A Novel Physics Informed Neural Network Framework for Solid State Phase Transformations
Accelerating Creep-Resistant Aluminum Alloy Design Through Generative AI-Driven Computational Models and Robust Validation
Accelerating Semiconductor Design with DFT and Deep Learning: Multiscale Modeling with Quantum Mechanical Accuracy
Adaptive Latent Space Tuning to Enable Characterizing Materials Dynamics Using Bragg Coherent Diffraction Imaging
Advanced Prediction of Crystalline Material Behavior Using Physics-Informed Neural Networks and Object-Oriented Crystal Plasticity Finite Element
Advancing Alloy Design: A Parameter-Free Statistical Approach to Predicting Stress Strain Curves of BCC Polycrystals
Advancing Automated Classification of Crystallographic Structures Using Synthetic Two-Dimensional X-Ray Diffraction Patterns and Deep Learning
AI-Powered Interface: Fully Automated Tool for LAMMPS Simulation and Analysis
AtomGPT: Atomistic Generative Pre-Trained Transformer for Forward and Inverse Materials Design
Bayesian Prediction and Optimization of Al-Ce-La-Nd-Mg-Ni Alloys’ Mechanical Properties Post Heat Treatment
Boosting Recyclable Plastic Sorting with AI-Generated Images and Vision Technology
Borides as Promising M2AX Phase Materials with High Elastic Modulus Using Machine Learning and Optimization
Charting the Large Chemical Space of Zintl Phases Using Graph Neural Networks
Closed-Loop Discovery of Lunar In Situ Resource Utilization (ISRU) Constrained Manufacturable Materials
Comparison of Phenomenological and Machine Learning Approaches to Model Inconel 718 Recrystallization Mechanisms
Computational Design and Optimization of 2D Spinodal Metallic Metamaterials for Improved Structural Behavior
Data-Driven Assessment and Selection of Magnetocaloric Materials by Grey Relational Analysis
Data Assimilation for Columnar Dendrite Growth Coupled with Large-Scale Phase-Field Simulations
Deep Gaussian Process Based Bayesian Optimization for Materials Discovery in High Entropy Alloy Space
Deep Learning-Based Image Denoising for Enhanced CT Image Reconstruction
Determination of Phase-Field Model Parameters Using Machine Learning Approach
Developing an Oxidation Materials Ontology for Data-Driven Materials Design
Digital Twin Application for Carbon Fiber Reinforced Polymer Composite Manufacturing
Discovery of Quaternary Structural Materials Using Constrained Generative Models
Enhancing Materials Discovery in Complex Composition Spaces: FUSE Meets Generative ML
From Pure Metals to Multi-Principal Element Alloys – A Physics-Informed Data-Driven Approach for Alloy Design
Generative Priors for Regularizing Ill-Posed Problems: Applications to 3D Polycrystalline RVE's
GrainPaint - A Multi-Scale Diffusion-Based Generative Model for Microstructure Reconstruction of Large-Scale Objects
High-Throughput Bayesian Calibration of Elastic-Plastic-Damage Model Parameters Using a Small Punch Test
Implications of X-CT Reliability to Determine Anomaly Distribution for Fatigue Assessment
Intelligent Optimization Algorithm-Based Optimization Model of Water Volume in Secondary Cooling Zone of Continuous Casting
Inverse Design of Architected Composite Materials with Desired Mechanical Behavior Based on Conditional Diffusion Model
J-1: A Data Structure and Collection System for Experimental Processes in Materials Science
J-2: Construction of Physical Property Temperature-Dependency Benchmarks for High-Accuracy Machine Learning Models
J-3: Data-Centric Engineering of Materials and Structures - Dissimilar Metal Welds as Use Case
J-4: Data Management of Through-Life Structural Integrity Assessment of Dissimilar Metal Welds for Nuclear Application
J-57: Data-Driven Study on Multi-Target Prediction of Mechanical Properties of Aluminum Alloys
J-5: Enhancing Computational Materials Research Through Large Language Model (LLM) Interfaces
J-6: Forward Prediction and Inverse Design of Additively Manufacturable Alloys via Autoregressive Language Models
J-7: Investigation of Phase-Field Data Assimilation System for Dendrite Growth Problem
J-8: ML-DiCE: A Machine Learning Framework for Predicting Diffusion Coefficients in Impure Metallic and Multi-Component Alloy Media
J-9: Optimizing Microstructure Prediction and Control in Advanced Structural Materials Using Deep Generative Models and Physics-Based Methods
KnowMat: Transforming Unstructured Material Science Literature into Structured Knowledge
Machine-Learning Structural Stability of Complex Intermetallic Phases
Machine Learning-Based Constitutive Model Parameter Estimation
Minimizing Hysteresis in NiTi-Based Non-Platinum Group Metal (PGM) High Temperature Shape Memory Alloys via Experimentally Validated Neural Network Machine Learning
Models of Magnetic Properties for Rapid Screening of Alternative Materials
Multi-Fidelity Deep Learning Approach for Designing Single-Phase BCC Refractory Multi-Principal Element Alloys (RMPEA) Across Various Temperatures
Navigating High-Dimensional Formulation Spaces with GP-Latent Variable Models for Dental Composites
Novel Superhard Materials Synthesis Using Generative ML
Physics-Based Priors for Phase Classification in Alloy Design
Physics-Informed Machine Learning of Thermal Stress Evolution in Laser Metal Deposition
Predicting Microstructure-Property Linkage in Alloys Using Graph Neural Network
Prediction of Nitrogen Content in Converter Based on an LLE-RF Model
Quantifying Error and Uncertainty in Transmission Electron Microscopy Images of Irradiation Defects
Quantifying Uncertainty of Object Detection Models in Electron Microscopy
Rapid Crystal Structure Prediction with the Aid of AI Generative Model and Descriptor Based Optimization
Representativeness of Experimentally Observed Microstructures for Validation and Uncertainty Propagation
Self-Supervised Learning (SSL) for Crystal Property Prediction via Structure Denoising
Sensitivity Analysis and Uncertainty Quantification in Process-Structure-Property of IN718
Structural Causal Learning with Atomistic Simulations for Advanced Materials
Structural Constraint Integration in Generative Model for Discovery of Quantum Material Candidates
Student’s T Process for Optimizing Material Properties
The Unsaturation Effect: Balanced Data Aggregation for Materials Informatics via Acquisition Functions
Uncertainty Quantification In Crystal Plasticity Simulations Using Multimodal High-Energy Synchrotron X-Ray Experiments
Utilization of Neural Networks and Numerical Modeling for Microstructural Analysis of Aluminum 6060 Alloy Composites
Verification, Validation, and Uncertainty Quantification for Self-Driving Labs - Proof of Concept with Chocolate as Frugal Twin
Questions about ProgramMaster? Contact programming@programmaster.org | TMS Privacy Policy | Accessibility Statement