Search Results for author:
Found 2 papers, 2 papers with code
What the Weight?! A Unified Framework for Zero-Shot Knowledge Composition
The knowledge encapsulated in a model is the core factor determining its final performance on downstream tasks.
ScaLearn: Simple and Highly Parameter-Efficient Task Transfer by Learning to Scale
We introduce ScaLearn, a simple and highly parameter-efficient two-stage MTL method that capitalizes on the knowledge of the source tasks by learning a minimal set of scaling parameters that enable effective knowledge transfer to a target task.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
