Research premise
This page reconstructs a plausible Sullivan work algorithm from public bibliographic evidence. It does not reproduce papers or books. A case is a thesis/result abstraction: an official publication, MIT-note section, book chapter, seminar unit, or lecture-style topic is classified by the methods it appears to deploy.
The corpus uses the Stony Brook publication list as its primary spine and extends it with Sullivan's MIT notes, string topology programme, CUNY seminar themes, and late fluid/computation papers. Because cases carry several strategy tags, prevalence percentages intentionally need not sum to 100%.
Interactive decision tree
Strategy prevalence ranking
Histogram width is count / 300 cases. These are overlapping method labels, not a probability simplex.
300-case corpus
| # | Year | Source | Case | Thesis / result abstraction | Strategy path |
|---|