Connect with us

Hi, what are you looking for?

CrazyFlux.comCrazyFlux.com

Tech News

Microsoft LASERs away LLM inaccuracies

Microsoft logo
Illustration: The Verge

During the January Microsoft Research Forum, Dipendra Misra, a senior researcher at Microsoft Research Lab NYC and AI Frontiers, explained how Layer-Selective Rank Reduction (or LASER) can make large language models more accurate.

With LASER, researchers can “intervene” and replace one weight matrix with an approximate smaller one. Weights are the contextual connections models make. The heavier the weight, the more the model relies on it. So, does replacing something with more correlations and contexts make the model less accurate? Based on their test results, the answer, surprisingly, is no.

“We are doing intervention using LASER on the LLM, so one would expect that the model loss should go up as we are doing more approximation,…

Continue reading…

You May Also Like

Editor's Pick

Thomas A. Berry and Nathaniel Lawson Chief Justice John Marshall famously stated in Marbury v. Madison (1803) that “[i]t is emphatically the province and duty...

Editor's Pick

Travis Fisher The Federal Energy Regulatory Commission (FERC) is not a household name, but its impact on the daily lives of Americans is profound....

Editor's Pick

Patrick G. Eddington If government bureaucrats are internally describing the civil society organization you work for as “vexsome,” it’s a good sign you and your...

World News

Scott R. Sehon tries to be intellectually honest in his critique of capitalism and his endorsement of socialism, but David Gordon writes that Sehon...