Marco Canini

distributed training (3) gradient compression (2) programmable switch (2) reinforcement learning (1) prompt engineering (1) chain-of-thought reasoning (1) performance optimization (1) neural network optimization (1) formal verification (1) communication compression (1) distributed machine learning (1) convergence rate (1) communication complexity (1) model training (1) black-box optimization (1) actor-critic method (1) prompt optimization (1) continuous action (1) instruction optimization (1) stochastic gradient (1)