Yi Mei

Runtime for complex project-scheduling simulations can be slashed from over 1,200 seconds to under 200 seconds using agentic AI optimizations, saving researchers significant computational resources.

Using a large language model as an external trainer, LaT boosts multi-task neural solvers' performance on diverse Vehicle Routing Problems without the computational burden of traditional meta-learning.

Forget hand-tuning: AutoSG automatically conjures high-performing solvers for expensive optimization problems directly from task prompts.