Praful contributed to the pytorch/pytorch repository by enhancing the Conv2d API documentation, focusing on clarifying the relationship between mathematical notation and API parameters. Using Python and leveraging his background in deep learning and machine learning, he updated the documentation to explicitly map variables such as C_in and C_out to their corresponding parameters, in_channels and out_channels. He also improved the reference for cross-correlation by linking to a more relevant section of Wikipedia, supporting clearer understanding for new contributors. The work addressed onboarding challenges and reduced ambiguity, demonstrating attention to detail and a thoughtful approach to technical documentation within the codebase.