Transformer Feed-Forward Layers Build Predictions by Promoting Concepts in the Vocabulary Space

Mor Geva, Avi Caciularu, Kevin Ro Wang, Yoav Goldberg

Research output: Contribution to conferencePaperpeer-review

24 Scopus citations

Fingerprint

Dive into the research topics of 'Transformer Feed-Forward Layers Build Predictions by Promoting Concepts in the Vocabulary Space'. Together they form a unique fingerprint.

Computer Science

Engineering

Keyphrases