storage efficiency

4 papers

• Compress & Cache: Vision token compression for efficient generation and retrieval
• HBLLM: Wavelet-Enhanced High-Fidelity 1-Bit Quantization for LLMs
• Object-X: Learning to Reconstruct Multi-Modal 3D Object Representations
• Temporal Smoothness-Aware Rate-Distortion Optimized 4D Gaussian Splatting