fix(joss_paper) solve conflict

Author: casenave

docs/joss_paper/paper.md

@@ -34,7 +34,7 @@ PLAID (Physics-Learning AI Datamodel) is a Python library and data format for re
 
 Machine learning for physical systems often suffers from inconsistent data representations across different domains and simulators.  Existing initiatives typically target narrow problems: e.g., separate formats for CFD or for finite-element data, and dedicated scripts to process each new dataset. This fragmentation hinders reproducibility and reuse of high-fidelity data.
 
-PLAID addresses this gap by providing a generic, unified datamodel that can describe many physics simulation data.  It leverages the CGNS standard [@poinot2018seven] to capture complex geometry and time evolution: for example, CGNS supports multi-block topologies and evolving meshes, with a data model that separates abstract topology (element families, etc.) from concrete mesh coordinates.  On top of CGNS, PLAID layers a lightweight organizational structure
+PLAID addresses this gap by providing a generic, unified datamodel that can describe many physics simulation data.  It leverages the CGNS standard [@poinot2018seven] to capture complex geometry and time evolution: for example, CGNS supports multi-block topologies and evolving meshes, with a data model that separates abstract topology (element families, etc.) from concrete mesh coordinates.  On top of CGNS, PLAID layers a lightweight organizational structure.
 
 By promoting a common standard, PLAID makes physics data interoperable across projects. It has already been used to package and publish multiple datasets covering structural mechanics and computational fluid dynamics. These PLAID-formatted datasets (hosted on Zenodo and Hugging Face) have supported ML benchmarks, democratizing access to simulation data.
 
@@ -52,7 +52,7 @@ By promoting a common standard, PLAID makes physics data interoperable across pr
 
 # Usage and Applications
 
-PLAID is designed for AI/ML researchers and practitioners working with simulation data. Various datasets, including 2D/3D fluid and structural simulations, are provided in PLAID format in [Hugging Face](https://huggingface.co/PLAID-datasets) and [Zenodo](https://zenodo.org/communities/plaid_datasets). Interactive benchmarks are hosted in a [Hugging Face community](https://huggingface.co/PLAIDcompetitions) on these datasets, providing detailed instructions and PLAID commands for data retrieval and manipulation (see [@casenave2025physics]). These datasets are also used in recent publications to illustrate the performance of the proposed scientific ML methods. In [@casenave2024mmgp; @kabalan2025elasticity; @kabalan2025ommgp],  Gaussian-process regression methods with mesh morphing are applied to these datasets. In [@perez2024gaussian; @perez2024learning] the datasets are leveraged in graph-kernel regression methods applied to fluid/solid mechanics.
+PLAID is designed for AI/ML researchers and practitioners working with simulation data. Various datasets, including 2D/3D fluid and structural simulations, are provided in PLAID format in [Hugging Face](https://huggingface.co/PLAID-datasets) and [Zenodo](https://zenodo.org/communities/plaid_datasets). Interactive benchmarks are hosted in a [Hugging Face community](https://huggingface.co/PLAIDcompetitions) on these datasets, providing detailed instructions and PLAID commands for data retrieval and manipulation, see [@casenave2025physics]. These datasets are also used in recent publications to illustrate the performance of the proposed scientific ML methods. In [@casenave2024mmgp; @kabalan2025elasticity; @kabalan2025ommgp],  Gaussian-process regression methods with mesh morphing are applied to these datasets. In [@perez2024gaussian; @perez2024learning] the datasets are leveraged in graph-kernel regression methods applied to fluid/solid mechanics.
 
 In summary, PLAID provides a comprehensive framework for physics-based ML data. By combining a unified data model, support for advanced mesh features, and helpful utilities, it addresses the need for interoperable, high-fidelity simulation datasets. Future enhancements involve developing general-purpose PyTorch dataloaders compatible with PLAID, along with establishing standardized evaluation metrics and unified pipelines for training and inference using the PLAID framework.