الانتقال إلى المحتوى الرئيسي

كتابة تعليق

مراجعة PREreview لPI-JEPA: Label-Free Surrogate Pretraining for Coupled Multiphysics Simulation via Operator-Split Latent Prediction

نُشرت
DOI
10.5281/zenodo.20534844
الترخيص
CC BY 4.0

This paper contains several hallucinated references, among which are the following.

  • S. Wang, M. Bharat, and P. Perdikaris. Improved architectures and training recipes for deep operator networks. arXiv preprint arXiv:2204.13678, 2022a

    • No such paper exists. The arxiv ID points to Ye, “Unified Simulation, Perception, and Generation of Human Behavior”

  • S. Wang and P. Perdikaris. Improved training of physics-informed neural networks with model ensembles. arXiv preprint arXiv:2204.05108, 2022.

    • The authors of this paper are Katsiaryna Haitsiukevich and Alexander Ilin.

  • M. De Hoop, D. Huang, W. Qian, and A. M. Stuart. Equivariant neural operators. arXiv preprint arXiv:2204.11139, 2022.

    • No such paper exists. The arxiv ID points to Mijangos et al., “Musical Stylistic Analysis: A Study of Intervallic Transition Graphs via Persistent Homology”

  • J. E. Santos, A. Mehrabifard, M. Prodanovi´c, and M. T. Balhoff. Hybrid deep neural operator/finite element method for unsaturated seepage flow. Advances in Water Resources, 195:104849, 2025.

    • No such paper exists. This appears to be a mangled version of He et al., “A hybrid deep neural operator/finite element method for ice-sheet modeling” (10.1016/j.jcp.2023.112428)

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they did not use generative AI to come up with new ideas for their review.

يمكنك كتابة تعليق على مراجعة PREreview هذه لـ PI-JEPA: Label-Free Surrogate Pretraining for Coupled Multiphysics Simulation via Operator-Split Latent Prediction.

قبل أن تبدأ

سنطلب منك تسجيل الدخول باستخدام ORCID iD الخاص بك. وإذا لم يكن لديك ORCID iD، يمكنك إنشاء واحد.

ما هو معرّف ORCID iD؟

ORCID iD هو معرّف فريد يميّزك عن جميع الأشخاص الذين يحملون الاسم نفسه أو اسمًا مشابهًا.

ابدأ الآن