Conference Logo ProgramMaster Logo
Conference Tools for 2027 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 2027 TMS Annual Meeting & Exhibition
Symposium Accelerating Innovation in Materials and Manufacturing
Presentation Title From Transferability to Prediction: The Error Geometry of Interatomic Potentials
Author(s) Alex S. Welcing
On-Site Speaker (Planned) Alex S. Welcing
Abstract Scope Every interatomic potential, from the embedded-atom method to foundation machine-learning models, is fit to limited data and asked to transfer beyond it. Transferability, where and why a potential fails, has been the field's recurring question, and the community has answered it institution by institution: OpenKIM, the NIST Interatomic Potentials Repository, growing benchmark suites. The argument here is that failure itself is now a measurable object, and that characterizing it demands the field's own evidentiary standards. Aggregating prediction error across roughly 900 published potentials by random-effects meta-analysis, and identifying its drivers by causal inference, reveals that error concentrates on a low-dimensional hyper-ribbon, the same sloppy-model geometry seen elsewhere in physics. That geometry survives the classical-to-machine-learning transition across fourteen of fifteen benchmark metals, making transferability a property one can predict rather than discover case by case. Doing this honestly means falsifiable claims and public self-correction, not curated success stories.
Proceedings Inclusion? Planned:
Keywords Machine Learning, Modeling and Simulation, Computational Materials Science & Engineering

OTHER PAPERS PLANNED FOR THIS SYMPOSIUM

A Concurrent Engineering Approach for Establishing a Multi-Planetary Society and  a Sustainable Energy Future
Accelerating Complex Materials Scale up
Accelerating Critical Mineral Recovery and Separation from Complex Domestic Resources
Accelerating Innovation in Materials Discovery and Manufacturing for Energy Technologies
Agentic Systems Design of an Open-Source Powder Doser for L-PBF Feedstock Research: CAD, PCB, and Firmware
Asset Intelligence for Belt Conveyor Systems: Accelerating Reliability, Digital Maintenance, and Operational Resilience in Mining
Automated, Self-Driving Materials Microscopy and Characterization Workflows
Automotive Body-In-White Material Selection: Early Stage Architectural Strategy
Autonomous Machine for Accelerated Structural Alloys Discovery and Manufacturing
Bridging Art and Engineering Through AI-Assisted Design and Structural Analysis of Metal Sculptures
Catalyzing Market Disruption: How DARPA's Innovation Model is Shaping the Future of Materials and Manufacturing
From Discovery to Deployment: Accelerating the Commercialization of Advanced Materials
From Transferability to Prediction: The Error Geometry of Interatomic Potentials
Genomic Materials Design: Making CyberSteels Fly
LEFFF: Accelerating HALEU Fuel Fabrication from Pilot Scale to Deployment
Leveraging AI & Materials Innovations to Create World-Changing Businesses
Perspectives on building new manufacturing companies
Scaling high-temperature alloy innovation for next-generation aerospace, defense, energy, and electrification markets
Structure-Property Relationships in AA1050 Aluminum Processed by Accumulative Roll Bonding: Effect of Reduction Ratio and Number of Passes.
Tensegrity-Inspired Lattices for Orientation-Invariant Impact Protection
The Old Man and the Factory: Short Stories from the Plant Floor
Translation of Aerospace Materials Research into Commercial Production
Ultrasonic Welding of Aluminum to Silicon

Questions about ProgramMaster? Contact programming@programmaster.org | TMS Privacy Policy | Accessibility Statement