Zhenmei Shi

Latest

• Conv-basis: A new paradigm for efficient attention inference and gradient computation in transformers (arXiv 2024)
• Exploring the frontiers of softmax: Provable optimization, applications in diffusion model, and beyond (arXiv 2024)
• Fast John Ellipsoid Computation with Differential Privacy Optimization (arXiv 2024)
• Fourier circuits in neural networks: Unlocking the potential of large language models in mathematical reasoning and modular arithmetic (arXiv 2024)
• Tensor attention training: Provably efficient learning of higher-order transformers (arXiv 2024)
• Toward Infinite-Long Prefix in Transformer (arXiv 2024)
• Unraveling the Smoothness Properties of Diffusion Models: A Gaussian Mixture Perspective (arXiv 2024)