Yifan Zhao
3 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
The paper proposes a disentangled representation framework to significantly improve few-shot layout-to-image generation by separating semantic identity from local visual details, thereby mitigating representation fragmentation.
The paper introduces TVIR, a new benchmark and multi-agent framework for deep research, to evaluate and improve the generation of factually reliable, text-visual interleaved reports.
The paper proposes PagedWeight, a method for managing Mixture-of-Experts (MoE) language model serving in KV-cache-intensive scenarios, achieving FP16-equivalent accuracy with up to 72.0% GPU memory savings and 1.94$ imes$ throughput improvement.
Papers
PagedWeight: Efficient MoE LLM Serving with Dynamic Quality-Aware Weight Quantization
The paper proposes PagedWeight, a method for managing Mixture-of-Experts (MoE) language model serving in KV-cache-intensive scenarios, achieving FP16-equivalent accuracy with up to 72.0% GPU memory sa…