Daniel Soudry

6 papers Technion

• Alias-Free ViT: Fractional Shift Invariance via Linear Attention
• Are Greedy Task Orderings Better Than Random in Continual Linear Regression?
• FP4 All the Way: Fully Quantized Training of Large Language Models
• Optimal Rates in Continual Linear Regression via Increasing Regularization
• Temperature is All You Need for Generalization in Langevin Dynamics and other Markov Processes
• Tensor-Parallelism with Partially Synchronized Activations