This paper reviews recent advances in implicit representation-based 3D shape generation. Implicit representations, such as signed distance fields (SDFs), radiance fields (RFs), and triplanes, have gained significant attention due to their ability to represent and generate 3D shapes effectively. SDFs provide inside-outside information for geometry reconstruction, while RFs combine geometry and appearance. Triplanes compress 3D data into 2D planes, enabling efficient generation. Various architectures, including GANs, VAEs, and diffusion models, have been applied to generate 3D shapes. Recent studies have shown that implicit representations can generate high-quality 3D shapes, but challenges remain in terms of generation speed, scalability, and integration with other representations. The paper categorizes recent works based on the type of representation and generation architecture, highlighting key contributions and open problems in the field. Future research directions include improving generation quality, increasing speed, and exploring larger-scale and multi-object scenes. The integration of implicit and explicit representations also holds promise for better editing and local modifications. Overall, implicit representations offer a flexible and powerful approach for 3D shape generation, but further research is needed to address current limitations.This paper reviews recent advances in implicit representation-based 3D shape generation. Implicit representations, such as signed distance fields (SDFs), radiance fields (RFs), and triplanes, have gained significant attention due to their ability to represent and generate 3D shapes effectively. SDFs provide inside-outside information for geometry reconstruction, while RFs combine geometry and appearance. Triplanes compress 3D data into 2D planes, enabling efficient generation. Various architectures, including GANs, VAEs, and diffusion models, have been applied to generate 3D shapes. Recent studies have shown that implicit representations can generate high-quality 3D shapes, but challenges remain in terms of generation speed, scalability, and integration with other representations. The paper categorizes recent works based on the type of representation and generation architecture, highlighting key contributions and open problems in the field. Future research directions include improving generation quality, increasing speed, and exploring larger-scale and multi-object scenes. The integration of implicit and explicit representations also holds promise for better editing and local modifications. Overall, implicit representations offer a flexible and powerful approach for 3D shape generation, but further research is needed to address current limitations.