Pruning LLMs Like a Physicist: Block Removal as an Ising Optimization Problem
中文摘要
该研究将大语言模型剪枝视为物理学中的伊辛优化问题，通过块移除方法并利用能量最小化原理来优化模型效率。
English Summary
This research applies Ising optimization from physics to prune LLMs via block removal, treating the task as an energy minimization problem to optimize efficiency.