{"paper":{"title":"Development of a 3D-CNN-based Prediction Model for Migration Barriers in Plasma-Wall Interactions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A 3D convolutional neural network predicts hydrogen migration barriers in tungsten to within 0.124 eV while running over 23000 times faster than the Nudged Elastic Band method.","cross_cats":["cond-mat.mtrl-sci"],"primary_cat":"physics.plasm-ph","authors_text":"Hiroaki Nakamura, Kazuo Hoshino, Keisuke Takeuchi, Seiki Saito, Shohei Yamoto, Yasuhiro Oda, Yuki Homma, Yuki Uchida","submitted_at":"2026-04-07T07:18:58Z","abstract_excerpt":"Understanding the long-term transport of hydrogen isotopes in plasma-facing materials, such as tungsten, is critical for the steady-state operation of magnetic confinement fusion reactors. However, dynamically updating the transition parameters for kinetic Monte Carlo (kMC) simulations as the atomic structure evolves under continuous plasma irradiation remains a severe computational bottleneck. Conventionally, calculating these migration barriers requires the iterative and computationally expensive Nudged Elastic Band (NEB) method. To overcome this limitation, this article presents a highly ef"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"the model demonstrated robust predictive accuracy, achieving a Mean Absolute Error (MAE) of 0.124 eV and a high coefficient of determination of 0.890. Furthermore, utilizing GPU acceleration, the inference time is reduced to approximately 2.7 milliseconds per barrier, achieving a speed-up ratio of over 23,000 compared to conventional NEB calculations.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That a 3D-CNN trained on static EAM-generated configurations will continue to give accurate barriers for the continuously evolving, non-equilibrium atomic structures that arise under sustained plasma irradiation, and that the two-channel volumetric input fully encodes all relevant environmental information.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A 3D-CNN surrogate predicts W-H migration barriers with 0.124 eV MAE and runs 23,000 times faster than NEB, enabling on-the-fly hybrid MD/kMC modeling of plasma-wall interactions.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A 3D convolutional neural network predicts hydrogen migration barriers in tungsten to within 0.124 eV while running over 23000 times faster than the Nudged Elastic Band method.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"2c07815eb7dd19fe54f5c9711ce5b3c80e5eb5d85daa3200e1660fbe506d0771"},"source":{"id":"2604.05521","kind":"arxiv","version":2},"verdict":{"id":"97515ad3-9e0d-4322-99f0-bea6fb428d0d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T18:56:34.383352Z","strongest_claim":"the model demonstrated robust predictive accuracy, achieving a Mean Absolute Error (MAE) of 0.124 eV and a high coefficient of determination of 0.890. Furthermore, utilizing GPU acceleration, the inference time is reduced to approximately 2.7 milliseconds per barrier, achieving a speed-up ratio of over 23,000 compared to conventional NEB calculations.","one_line_summary":"A 3D-CNN surrogate predicts W-H migration barriers with 0.124 eV MAE and runs 23,000 times faster than NEB, enabling on-the-fly hybrid MD/kMC modeling of plasma-wall interactions.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That a 3D-CNN trained on static EAM-generated configurations will continue to give accurate barriers for the continuously evolving, non-equilibrium atomic structures that arise under sustained plasma irradiation, and that the two-channel volumetric input fully encodes all relevant environmental information.","pith_extraction_headline":"A 3D convolutional neural network predicts hydrogen migration barriers in tungsten to within 0.124 eV while running over 23000 times faster than the Nudged Elastic Band method."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.05521/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"3015ea0a243866f14efe2dbdec60bf8e68ab4566ca968aa17fa99ddcfe3fd473"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}