output quality

5 papers

• Cue3D: Quantifying the Role of Image Cues in Single-Image 3D Generation
• DynamicRAG: Leveraging Outputs of Large Language Model as Feedback for Dynamic Reranking in Retrieval-Augmented Generation
• ESCA: Enabling Seamless Codec Avatar Execution through Algorithm and Hardware Co-Optimization for Virtual Reality
• LinEAS: End-to-end Learning of Activation Steering with a Distributional Loss
• Policy Optimized Text-to-Image Pipeline Design