GitHub

Optimizing Large Language Models Using Layer-Selective Rank Reduction and Random Matrix Theory

In this study, we introduce a novel adaptation of Layer-Selective Rank Reduction (LASER) for optimizing large language models, utilizing Marchenko-Pastur law from Random Matrix Theory. This approach ...
Microsoft

Improving Reasoning in Language Models with LASER: Layer-Selective Rank Reduction

“An LLM is trained on lots of data, often collected from the internet, and uses a model architecture, typically a transformer, to train the model, and they work remarkably well across a range of ...
IEEE

Simultaneous denoising and reconstruction of 5-D seismic data via damped rank-reduction method

Abstract: The Cadzow rank-reduction method can be effectively utilized in simultaneously denoising and reconstructing 5-D seismic data that depend on four spatial dimensions. The classic version of ...