The field of 3D Gaussian Splatting is moving towards improving rendering quality, efficiency, and scalability. Researchers are exploring new methods to reduce storage costs, accelerate rendering, and enhance geometric reconstruction. Notable advancements include the development of self-adaptive compression methods, proxy-guided weighted blending, and reinforcement learning-based hyperparameter tuning. These innovations have led to significant improvements in rendering quality, speed, and memory efficiency. Noteworthy papers include SA-3DGS, which achieves up to 66x compression while maintaining rendering quality. Duplex-GS proposes a dual-hierarchy framework that integrates proxy Gaussian representations with order-independent rendering techniques, achieving photorealistic results while sustaining real-time performance. RLGS introduces a reinforcement learning framework for adaptive hyperparameter tuning, consistently enhancing rendering quality across multiple state-of-the-art 3DGS variants. DET-GS proposes a unified depth and edge-aware regularization framework, significantly enhancing structural fidelity and robustness against depth estimation noise. Perceive-Sample-Compress introduces a novel framework for real-time 3D Gaussian Splatting, improving memory efficiency and high visual quality while maintaining real-time rendering speed. Refining Gaussian Splatting proposes a volumetric densification approach, surpassing 3DGS in reconstruction quality. 3DGabSplat leverages a novel 3D Gabor-based primitive for radiance field representation, achieving state-of-the-art rendering quality across both real-world and synthetic scenes.