Papers with domain expert mixture

1 papers
DEMix Layers: Disentangling Domains for Modular Language Modeling (2022.naacl-main)

Copied to clipboard

Challenge: Extensive experiments with autoregressive transformer LMs show that DEMix layers reduce test-time perplexity and increase training efficiency.
Approach: They introduce a new domain expert mixture layer that enables conditioning a language model on the domain of the input text.
Outcome: Experiments with 1.3B LMs show that DEMix layers reduce test-time perplexity, increase training efficiency, and enable rapid adaptation.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations