Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned
2019 · 1,008 citations
https://doi.org/10.18653/v1/p19-1580
Claim or correct this author profile
Claim this profile or suggest corrections to the name, affiliation, bio, photo, or paper titles. Approved changes appear as verified CitedEvidence overlays.
26
Papers
2.0K
Citations
14
h-index
14
i10-index
David Talbot is an academic researcher from Yandex School of Data Analysis. The author has contributed to research in topics: Natural Language Processing Techniques & Topic Modeling & Caching and Content Delivery. The author has an h-index of 14, co-authored 23 publications.
Click to start Chat