Deepseek

1. Physics-Informed Neural Networks (PINNs): A Deep Learning Framework for

Solving Forward and Inverse Problems Involving Nonlinear Partial Differential
Equations
Authors: Raissi, M., Perdikaris, P., Karniadakis, G.E.
Year: 2019
Journal: Journal of Computational Physics
Key Contribution: Introduced the core PINN framework, integrating physical laws
(via PDE residuals) into neural network loss functions for solving forward and
inverse problems.
Citation: @Raissi2019b
Matches title+abstract, body: 285; 16.23

2. Physics Informed Deep Learning (Part I and II)

Authors: Raissi, M., Perdikaris, P., Karniadakis, G.E.
Year: 2017–2018
Platform: arXiv
Key Contribution: Early foundational work demonstrating how neural networks
can encode physical principles, paving the way for modern PINNs.
Citation: @Raissi2017a, @Raissi2017b
Matches title+abstract, body: (a) 340; 11.14. (b) 230; 8.53

3. DeepXDE: A Deep Learning Library for Solving Differential Equations

Authors: Lu, L., Meng, X., Mao, Z., Karniadakis, G.E.
Year: 2021
Journal: SIAM Review
Key Contribution: Introduced a versatile Python library for implementing PINNs
and solving diverse PDEs, enhancing accessibility and reproducibility.
Citation: @LuLu2019
Matches title+abstract, body: 151; 44.0

4. Artificial Neural Networks for Solving Ordinary and Partial Differential

Equations
Authors: Lagaris, I.E., Likas, A., Fotiadis, D.I.
Year: 1998
Journal: IEEE Transactions on Neural Networks
Key Contribution: Early work using neural networks to solve differential
equations, often cited as a precursor to PINNs.
Citation: @Lagaris1997
Matches title+abstract, body: 170; 14.71

5. DGM: A Deep Learning Algorithm for Solving Partial Differential Equations

Authors: Sirignano, J., Spiliopoulos, K.
Year: 2018
Journal: SIAM Journal on Scientific Computing
Key Contribution: Proposed the Deep Galerkin Method (DGM), a related approach
using neural networks to solve high-dimensional PDEs.
Citation: @Sirignano2018
Matches title+abstract, body: 202; 14.32

6. Physics-Informed Machine Learning

Authors: Karniadakis, G.E., Kevrekidis, I.G., Lu, L., Perdikaris, P., Wang, S., Yang, L.
Year: 2021
Journal: Nature Reviews Physics
Key Contribution: A comprehensive review of PINNs and broader physics-
informed machine learning, highlighting applications and challenges.
Citation: @Karniadakis2021
Matches title+abstract, body: 149; 53.37

7. Adversarial Uncertainty Quantification in Physics-Informed Neural Networks

Authors: Yang, L., Meng, X., Karniadakis, G.E.
Year: 2021
Journal: Journal of Computational Physics
Key Contribution: Introduced adversarial training to improve PINN robustness
and uncertainty quantification.
Citation: @YangY2019
Matches title+abstract, body: 137; 12.91

8. PPINN: Parareal Physics-Informed Neural Networks for Time-Dependent

PDEs
Authors: Zhang, D., Guo, L., Karniadakis, G.E.
Year: 2020
Journal: Computer Methods in Applied Mechanics and Engineering
Key Contribution: Combined parallel-in-time algorithms with PINNs for efficient
long-time integration of PDEs.
Citation: @Meng2019
Matches title+abstract, body: 203; 31.59

9. On the Convergence of Physics Informed Neural Networks for Linear Second-

Order Elliptic and Parabolic Type PDEs
Authors: Shin, Y., Darbon, J., Karniadakis, G.E.
Year: 2020
Journal: Communications in Computational Physics
Key Contribution: Provided theoretical analysis of PINN convergence for specific
PDE classes.
Citation: @Shin2020
Matches title+abstract, body: 244; 14.45

10. A Hybrid Neural Network/Finite Difference Approach for Solving Inverse

Problems
Authors: Psichogios, D.C., Ungar, L.H.
Year: 1992
Journal: Computers & Chemical Engineering
Key Contribution: Early hybrid approach combining neural networks with
physics-based models, inspiring later PINN methodologies.
Citation: @Psichogios1992
Matches title+abstract, body: 23; 38.46

Foundational Works
1. Raissi, M., Perdikaris, P., & Karniadakis, G. E. (2019) @Raissi2019b
Physics-Informed Neural Networks: A Deep Learning Framework for Solving
Forward and Inverse Problems Involving Nonlinear Partial Differential
Equations
Journal of Computational Physics
Introduces the PINN framework for solving PDEs and inverse problems.
Matches title+abstract, body: 285; 16.23
2. Raissi, M., et al. (2017) @Raissi2017a
Physics Informed Deep Learning (Part I): Data-Driven Solutions of Nonlinear
PDEs
Early work on integrating physical laws into neural networks.
Matches title+abstract, body: 340; 11.14
3. Raissi, M., et al. (2017) @Raissi2017b
Physics Informed Deep Learning (Part II): Data-Driven Discovery of Nonlinear
 PDEs Focuses on inverse problems and parameter estimation.
Matches title+abstract, body: 230; 8.53
Methodological Advances
4. Jagtap, A. D., & Karniadakis, G. E. (2020) @Jagtap2020b
Adaptive Activation Functions Accelerate Convergence in Deep and Physics-
Informed Neural Networks
Proposes adaptive activation functions to improve training.
Matches title+abstract, body: 265; 15.54
5. Shukla, K., et al. (2020) @Shukla2021
Conservative Physics-Informed Neural Networks on Discrete Domains for
Conservation Laws
Introduces domain-decomposed cPINNs.
Matches title+abstract, body: 125; 28.83
6. Jagtap, A. D., et al. (2020) @Jagtap2020
Extended Physics-Informed Neural Networks (XPINNs): A Generalized Space-
Time Domain Decomposition
Extends cPINNs for complex geometries.
Matches title+abstract, body: 371; 49.17
7. Wang, S., et al. (2020) @Wang2020b
Understanding and Mitigating Gradient Pathologies in Physics-Informed Neural
Networks
Addresses spectral bias via Fourier feature networks.
Matches title+abstract, body: 233; 28.07
8. Kharazmi, E., et al. (2019) @Kharazmi2019
Variational Physics-Informed Neural Networks for Solving Partial Differential
Equations
Combines variational methods with PINNs (VPINNs).
Matches title+abstract, body: 389; 15.04
9. Yang, L., et al. (2020) @YangLiu2021
B-PINNs: Bayesian Physics-Informed Neural Networks for Noisy Data
 Integrates Bayesian inference with PINNs.
Matches title+abstract, body: 278; 14.25
Applications
10. Hennigh, W., et al. (2021) @Hennigh2021
NVIDIA SimNet: An AI-Accelerated Multi-Physics Simulation Framework
Industrial application to fluid dynamics.
Matches title+abstract, body: 55; 13.5
11. Sahli Costabal, F., et al. (2020) @LuLu2019
Physics-Informed Neural Networks for Cardiac Activation Mapping
Applies PINNs to cardiac electrophysiology.
Matches title+abstract, body: 151; 44
12. Cai, S., et al. (2021) @Cai2021c
Physics-Informed Neural Networks for Heat Transfer Problems
Solves heat transfer and phase change problems.
Matches title+abstract, body: 89; 43.75
13. Jin, X., et al. (2021) @Jin2021
NSFnets: Physics-Informed Neural Networks for the Navier-Stokes Equations
Focuses on incompressible fluid flows.
Matches title+abstract, body: 279; 8.27

Theoretical and Computational

Improvements
14. Lu, L., et al. (2021) @LuLu2019
DeepXDE: A Deep Learning Library for Solving Differential Equations
Introduces a widely used PINN library.
Matches title+abstract, body: 151; 44.0
15. Zhang, D., et al. (2019) @ZhangD2020
Solving Stochastic Differential Equations with PINNs
Extends PINNs to stochastic PDEs.
Matches title+abstract, body: 272; 11.58
16. Yeonjong Shin, Jerome Darbon, George Em Karniadakis @Shin2020
On the convergence of physics informed neural networks for linear second-order
elliptic and parabolic type PDEs
Theoretical analysis of PINN convergence.
Matches title+abstract, body: 244; 14.45
17. Krishnapriyan, A., et al. (2021) x (made up)
Estimating Convergence of Deep PDE Solvers Without Ground Truth
Analyzes error estimation in PINNs.
Matches title+abstract, body: N/A
Inverse Problems and Extensions
18. Alexandre M. Tartakovsky, Carlos Ortiz Marrero, Paris Perdikaris, Guzel D.
Tartakovsky, David Barajas-Solano @Tartakovsky2018
Learning Parameters and Constitutive Relationships with Physics Informed
Deep Neural Networks
Solves material modeling inverse problems.
Matches title+abstract, body: 185; 10.77
19. Chen, Y., et al. (2020) @Chen2020
PINNs for Inverse Problems in Nano-Optics and Metamaterials
Applies PINNs to photonics and metamaterials.
Matches title+abstract, body: 104; 23.0
20. Penwarden, M., et al. (2021) @Penwarden2021
Multi-Fidelity Physics-Constrained Neural Networks
Combines multi-fidelity data with PINNs.
Matches title+abstract, body: 103; 29.53

ChatGPT
1. Physics-Informed Neural Networks: A Deep Learning Framework for Solving
Forward and Inverse Problems Involving Nonlinear Partial Differential
Equations
Authors: Maziar Raissi, Paris Perdikaris, George Em Karniadakis
 Journal: Journal of Computational Physics, 2019
Summary: This foundational paper introduces the concept of PINNs,
demonstrating their capability to solve both forward and inverse problems
governed by PDEs.
Citation: @Raissi2019b
Matches title+abstract, body: 285; 16.23
2. The Deep Ritz Method: A Deep Learning-Based Numerical Algorithm for
Solving Variational Problems
Authors: Weinan E, Bing Yu
Journal: Communications in Mathematics and Statistics, 2018
Summary: This work presents the Deep Ritz method, leveraging deep learning
for numerical solutions of variational problems, closely related to the PINN
framework.
Citation: @Weinan2018
Matches title+abstract, body: 15; 4
3. DGM: A Deep Learning Algorithm for Solving Partial Differential Equations
Authors: Justin Sirignano, Konstantinos Spiliopoulos
Journal: Journal of Computational Physics, 2018
Summary: The authors propose the Deep Galerkin Method (DGM), a deep
learning approach for solving high-dimensional PDEs, laying groundwork
relevant to PINNs.
Citation: @Sirignano2018
Matches title+abstract, body: 202; 14.32
4. DeepXDE: A Deep Learning Library for Solving Differential Equations
Authors: Lu Lu, Xuhui Meng, Zhiping Mao, George Em Karniadakis
Journal: SIAM Review, 2021
Summary: This paper introduces DeepXDE, a comprehensive library designed
for solving differential equations using deep learning, facilitating the
implementation of PINNs.
Citation: @LuLu2019
Matches title+abstract, body: 151; 44
5. Universal Differential Equations for Scientific Machine Learning
Authors: Christopher Rackauckas et al.
Journal: arXiv preprint arXiv:2001.04385, 2020
Summary: The study discusses universal differential equations, combining
mechanistic and data-driven models, enhancing the flexibility of PINNs in
scientific machine learning.
Citation: @Rackauckas2021
Matches title+abstract, body: 71; 9.85
6. NeuralPDE: Automating Physics-Informed Neural Networks (PINNs) with
Error Approximations
Authors: Kirill Zubov et al.
Journal: arXiv preprint arXiv:2107.09443, 2021
Summary: This work presents NeuralPDE, an automated approach to
implementing PINNs with error approximations, improving their accuracy and
reliability.
Citation: @Zubov2021
Matches title+abstract, body: 135; 8.57
7. Physics-Informed Neural Networks for High-Speed Flows
Authors: Zhiping Mao, Ameya D. Jagtap, George Em Karniadakis
Journal: Computer Methods in Applied Mechanics and Engineering, 2020
Summary: The paper explores the application of PINNs to model high-speed
aerodynamic flows, addressing challenges in capturing shock waves and
discontinuities.
Citation: @Mao2020
Matches title+abstract, body: 137; 8.74
8. Hidden Fluid Mechanics: Learning Velocity and Pressure Fields from Flow
Visualizations
Authors: Maziar Raissi, Alireza Yazdani, George Em Karniadakis
Journal: Science, 2020
Summary: This study demonstrates the use of PINNs to infer hidden quantities
in fluid mechanics, such as velocity and pressure fields, from visual data.
 Citation: @Raissi2020
Matches title+abstract, body: 59; 6.56
9. Physics-Informed Neural Networks for Inverse Problems in Supersonic
Flows
Authors: Ameya D. Jagtap, Zhiping Mao, Nikolaus Adams, George Em
Karniadakis
Journal: Journal of Computational Physics, 2022
Summary: The authors apply PINNs to inverse problems in supersonic flows,
showcasing their potential in reconstructing flow properties from limited
data.
Citation: @Jagtap2022
Matches title+abstract, body: 118; 32.11
10. B-PINNs: Bayesian Physics-Informed Neural Networks for Forward and
Inverse PDE Problems with Noisy Data
Authors: Liu Yang, Xuhui Meng, George Em Karniadakis
Journal: Journal of Computational Physics, 2021
Summary: This paper introduces a Bayesian approach to PINNs, enhancing
their robustness in handling noisy data for both forward and inverse PDE
problems.
Citation: @YangLiu2021
Matches title+abstract, body: 278; 14.25
11. An Expert's Guide to Training Physics-Informed Neural Networks
Authors: Sifan Wang, Shyam Sankaran, Hanwen Wang, Paris Perdikaris
Journal: arXiv preprint arXiv:2308.08468, 2023
Summary: The authors provide best practices and benchmark problems to
improve the training efficiency and accuracy of PINNs.
Citation: @WangS2023
Matches title+abstract, body: 125; 21.72
12. Number Theoretic Accelerated Learning of Physics-Informed Neural
Networks
Authors: Takashi Matsubara, Takaharu Yaguchi
Journal: arXiv preprint arXiv:2307.13869, 2023
 Summary: This work introduces number theoretic methods to accelerate the
training of PINNs, reducing computational costs while maintaining accuracy.
Citation: @Matsubara2023
Matches title+abstract, body: 83; 29.5
13. PINNsFormer: A Transformer-Based Framework for Physics-Informed
Neural Networks
Authors: Zhiyuan Zhao, Xueying Ding, B. Aditya Prakash
Journal: arXiv preprint arXiv:2307.11833, 2023
Summary: The paper proposes a transformer-based architecture for PINNs,
aiming to capture temporal dependencies more effectively in time-dependent
PDEs.
Citation: @Zhao2023
Matches title+abstract, body: 197; 49.35
14. Physics-Informed Kolmogorov-Arnold Neural Networks for Dynamical
Analysis via Efficient-KAN and WAV-KAN
Authors: Subhajit Patra et al.
Journal: arXiv preprint arXiv:2407.18373,
Citation: @Patra2024
Matches title+abstract, body: 195; 13.94
15. Gradient-enhanced physics-informed neural networks for forward and
inverse PDE problems
Authors: Jeremy Yu, Lu Lu, Xuhui Meng, George Em Karniadakis
Journal: Computer Methods in Applied Mechanics and Engineering, 2021
Summary: .
Citation: @Yu2021
Matches title+abstract, body: 223; 23.5
16. Variational Physics-Informed Neural Networks for Solving Partial
Differential Equations
Authors: E. Kharazmi, Z. Zhang, G. E. Karniadakis
Journal: Computer Methods in Applied Mechanics and Engineering, 2020
Summary: The study develops a variational version of PINNs, leveraging
energy-based formulations to enhance convergence and stability.
 Citation: @Kharazmi2019
Matches title+abstract, body: 389; 15.04
17. hp-VPINNs: Physics-Informed Neural Networks with Domain Decomposition
Authors: Ehsan Kharazmi, Zhongqiang Zhang, George E.M. Karniadakis
Journal: Computer Methods in Applied Mechanics and Engineering Volume
374, 1 February 2021, 113547
Summary:
Citation: @Kharazmi2021
Matches title+abstract, body: 89; 16.24
18. Self-Adaptive Physics-Informed Neural Networks for Solving PDEs
Authors: Levi McClenny, Ulisses Braga-Neto
Journal:
Summary:
Citation: @McClenny2020
Matches title+abstract, body: 221; 18.96
19. # Multifidelity Modeling for Physics-Informed Neural Networks (PINNs)
Authors: Michael Penwarden, Shandian Zhe, Akil Narayan, Robert M. Kirby
Journal:
Summary:
Citation: @Penwarden2021
Matches title+abstract, body: 103; 29.53
20. Extended Physics-Informed Neural Networks (XPINNs): A Generalized
Space-Time Domain Decomposition Approach
Authors: Ameya D. Jagtap, Ehsan Kharazmi, George Em Karniadakis
Journal: SIAM Journal on Scientific Computing, 2020
Summary: XPINNs extend PINNs by introducing space-time domain
decomposition, allowing for localized training and enhanced performance in
large-scale problems.
Citation: @Jagtap2020
Matches title+abstract, body: 371; 49.17
Explanation of the numbers Matches title+abstract, body.
The first number is the number of matches of the title plus the abstract to a
set of keywords related to the objective of the study. The second number is
the number of matches to a second longer set of keywords divided by the
number of pages of the paper. The numbers are indication of relevance but it
is very subjective because of the selection of the keywords.