Yangyang Tan developed CELU activation function support for the NVIDIA/physicsnemo repository, expanding the library’s activation options for neural simulations. Using Python and deep learning techniques, Yangyang updated the ACT2FN mapping to include the 'celu' key, which references the torch.nn.CELU class with a default alpha of 1.0. This integration allows researchers to seamlessly select CELU as an activation function, improving model flexibility and reproducibility within the framework. The work included comprehensive tests to ensure correct behavior of the new activation path. Over the month, Yangyang focused on model development, delivering a targeted, well-integrated feature without introducing new bugs.