Abstract Transformers
Abstract Transformers is an abstract interpretation framework that generalizes transformer models by formalizing their computation as abstract domains and transformers over symbolic representations, enabling sound approximation of attention mechanisms in program analysis.
21
Publications
1.3K
Citations
37
Authors
26
Institutions
Publications per year
2017–2026
8
Parents
21
An abstract domain for certifying neural networks
Gagandeep Singh, Timon Gehr, Markus Püschel et al. · Proceedings of the ACM on Programming Languages · 2019 · 598 citations · Full text
Generating precise and concise procedure summaries
Greta Yorsh, Eran Yahav, Satish Chandra · 2008 · 68 citations
Automatic modular abstractions for linear constraints
David Monniaux · 2009 · 56 citations · Full text
1–5 of 21
37
3
h-index
3
h-index
3
h-index
3
h-index
3
h-index
1–5 of 37
26

Centre National de la Recherche Scientifique
Paris, France
3 h-index
3 publications
107 citations
ETH Zurich
Zurich, Switzerland
3 h-index
11 publications
2.5K citations
University of Wisconsin–Madison
Madison, United States
2 h-index
7 publications
200 citations

Tel Aviv University
Tel Aviv, Israel
2 h-index
2 publications
80 citations
IBM (United States)
Armonk, United States
2 h-index
4 publications
160 citations
Ludwig-Maximilians-Universität München
Munich, Germany
1 h-index
3 publications
483 citations

Université Grenoble Alpes
Saint-Martin-d'Hères, France
1 h-index
1 publications
31 citations
University of Washington
Seattle, United States
1 h-index
1 publications
19 citations
1–8 of 26
PACMPL · 4 publications · 658 citations
4
h-index
2 publications · 32 citations
2
h-index
2 publications · 32 citations
2
h-index
1 publications · 32 citations
1
h-index
TACL · 1 publications · 19 citations
1
h-index
1–5