Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/William Merrill

William Merrill

2 indexed papers

Recent (6 mo)
2
With code
0
Influential cites
0
Benchmarked
0

Publications per year

2
26

Top categories

ML×2AI×2Complexity×2NLP×2Formal Lang.×2

Frequent co-authors

Anej Svete2×
Ryan Cotterell2×
Ashish Sabharwal2×

Research Timeline

2026
Revisiting Padded Transformer Expressivity: Which Architectural Choices Matter and Which Don't

The paper analyzes the expressivity of padded transformers, proving that their computational power is primarily determined by model depth and numeric precision, rather than attention type or width.

Revisiting Padded Transformer Expressivity: Which Architectural Choices Matter and Which Don't

The paper analyzes the expressivity of padded transformers, proving that their computational power is primarily determined by model depth and numeric precision, rather than attention type or width.

Highlighted terms show continued research focus across papers

Papers

cs.LGcs.AIcs.CCRecentMay 28, 2026

Revisiting Padded Transformer Expressivity: Which Architectural Choices Matter and Which Don't

Anej Svete, William Merrill, Ryan Cotterell, Ashish Sabharwal

The paper analyzes the expressivity of padded transformers, proving that their computational power is primarily determined by model depth and numeric precision, rather than attention type or width.

View →
cs.LGcs.AIcs.CCRecentMay 28, 2026

Revisiting Padded Transformer Expressivity: Which Architectural Choices Matter and Which Don't

Anej Svete, William Merrill, Ryan Cotterell, Ashish Sabharwal

The paper analyzes the expressivity of padded transformers, proving that their computational power is primarily determined by model depth and numeric precision, rather than attention type or width.

View →