3 ms·

Having a PhD in math, I must say this is really wicked. Parsing through it, it looks legit to me. Proper typesetting, proper definition, theorem, proof structur

Having a PhD in math, I must say this is really wicked. Parsing through it, it looks legit to me. Proper typesetting, proper definition, theorem, proof structure. The usual phrases like "it is well known, that..." are used, the sentences make sense at a first glance. The only thing which looks suspicious are the formulas. They are too heavy and make no sense, as no explanation of the symbols are given. But kudos, this is really really good. Wouldn't be surprised if you can trick yourself in some low-tier journal using such a paper as template.

twangist 8y ago
Oh, hardly. This is nothing like good, in the sense of "successful forgery", let alone "really, really good". Did you write the program? At a distance of 10 feet, the document the program produces seems to nail the form and appearance of a math paper, probably because its template is a real paper. The formulas exhibited are or might be well-formed, but even at 5 feet away, it's clear that you've never see expressions that use so many obscure LaTeX glyphs. But try to read it, and it's immediately clear that no two sentences in a row even make use the same terms, and that every single sentence is sheer gibberish. The program is quite good at generating fake-math terminology, and knows more than a bit about naming conventions (e.g. coming up with "Perelman" as an adjective). But if, for example, this definition passes muster with you: > Definition 5.1. Let u_{l,V} be a Noetherian algebra. A homomorphism is a homomorphism if it is elliptic. then I wouldn't want to read your thesis either. Addendum: A for effort :) but F for the hype.
- blastbeat 8y ago
  I didn't write the program, and I only looked from a distance of 10 feet. On the other hand, I've seen my share of papers with "obscure LaTeX glyphs". So yes, it is really really good, and I'm undecided whether this project is a serious attempt of forgery or just malicious satire.
  - twangist 8y ago
    I hope merely satire. I don't find it convincing at all, so I wouldn't suspect malice, nor of course a serious attempt at forgery :)

twangist 8y ago

Oh, hardly. This is nothing like good, in the sense of "successful forgery", let alone "really, really good". Did you write the program? At a distance of 10 feet, the document the program produces seems to nail the form and appearance of a math paper, probably because its template is a real paper. The formulas exhibited are or might be well-formed, but even at 5 feet away, it's clear that you've never see expressions that use so many obscure LaTeX glyphs. But try to read it, and it's immediately clear that no two sentences in a row even make use the same terms, and that every single sentence is sheer gibberish. The program is quite good at generating fake-math terminology, and knows more than a bit about naming conventions (e.g. coming up with "Perelman" as an adjective). But if, for example, this definition passes muster with you: > Definition 5.1. Let u_{l,V} be a Noetherian algebra. A homomorphism is a homomorphism if it is elliptic. then I wouldn't want to read your thesis either. Addendum: A for effort :) but F for the hype.

blastbeat 8y ago
I didn't write the program, and I only looked from a distance of 10 feet. On the other hand, I've seen my share of papers with "obscure LaTeX glyphs". So yes, it is really really good, and I'm undecided whether this project is a serious attempt of forgery or just malicious satire.
- twangist 8y ago
  I hope merely satire. I don't find it convincing at all, so I wouldn't suspect malice, nor of course a serious attempt at forgery :)