Developed the Upsample Interpolation Mode Extension for the ml-explore/mlx repository, focusing on enhancing the Upsample layer by introducing cubic interpolation support. This work expanded the API to offer more granular control over interpolation methods, directly improving the flexibility and quality of downstream machine learning pipelines. The implementation emphasized robust library development practices in Python, with careful attention to type hinting for the mode parameter to ensure type safety and maintainability. All changes were thoroughly documented and linked to relevant issues and pull requests, demonstrating disciplined engineering and alignment with repository standards while preparing the feature for broader adoption.