Sharpened lower bounds for cut elimination (Q2892684)
From MaRDI portal
| This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: Sharpened lower bounds for cut elimination |
scientific article; zbMATH DE number 6047783
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Sharpened lower bounds for cut elimination |
scientific article; zbMATH DE number 6047783 |
Statements
Sharpened lower bounds for cut elimination (English)
0 references
19 June 2012
0 references
cut elimination
0 references
lower bounds
0 references
0.8935938
0 references
0 references
0.8858892
0 references
0 references
0.8807249
0 references
0.87852025
0 references
0.8781743
0 references
0.87732273
0 references
A previous lower bound for cut elimination was of the form \(2^{h(P)}_{\epsilon d}\), where \(h(P)\) is the height of the proof \(P\), \(d\) is the maximal complexity of cuts in \(P\) and \(\epsilon\sim 1/2\). The present paper establishes \(\epsilon\sim 1\). More precisely, \(\epsilon d\) is replaced by \(d-c\) where \(c\leq 8\). This is very close to the existing upper bounds for cut elimination from arbitrary proofs leaving a gap of only \(\log^*\), the inverse superexponential function. The argument in the present paper starts with short proofs with cuts (due to V. Orevkov and other authors) of the totality of \(2^n_x\) for every \(n\). Then inductive formulas are simplified in an equivalent way to lower their complexity.
0 references