Bethe Ansatz Solutions Using Large Language Models
Explore how Large Language Models solve Bethe Ansatz problems in integrable spin chains, advancing AI applications in theoretical physics.
Efficient Image Compression for Wind Turbine Inspections
Discover a deep learning framework that enhances wind turbine image compression using segmentation and dual-mode coding for accurate defect detection.
Deep Learning MRI Reconstruction Using Synthetic Fractals
Explore deep learning MRI reconstruction trained on synthetic fractal data, offering a privacy-safe alternative to traditional cardiac MRI datasets.
SISA: Efficient GEMM Acceleration with Scale-In Systolic Array
Discover SISA, a novel scale-in systolic array architecture that boosts GEMM performance by 8.5x and cuts energy-delay product by 93% for AI workloads.
UniRank: Domain-Specific Hybrid Text-Image Reranking
UniRank improves domain-specific multimodal reranking by efficiently scoring hybrid text-image candidates with vision-language models.