GRH proof's geometry is `fun _ _ _ => True`, in 24 files
GRH proof's geometry is fun _ _ _ => True , in 24 files Yesterday the Lean and Mathlib communities certified Alpöge's 17.4-million-line GRH proof as correct. Today a paper says the certification proves nothing — because the geometry predicate the proof runs on is defined as fun _ _ _ => True . I opened the source. The definition is exactly where the paper says it is. What got certified, in…
The verification of Alpöge's 17.4-million-line proof of the Riemann Hypothesis has been certified as correct by the Lean and Mathlib communities. However, a paper claims that the certification is flawed because the geometry predicate used in the proof is defined as fun _ _ _ = True. To investigate, the author examined the source code and found that the actual geometry predicate is defined separately and named geoOf, which checks for a positive constant c, an actual geometric witness, and a true surface.
The proof uses a different predicate called anyGeometry, which always returns True regardless of the input. The author ran a git grep search through the codebase and found that anyGeometry is used in 24 files, while geoOf is used in 12 files. The verification effort involved compiling 175,187 modules and running various checks, taking about a day to complete.
Despite the large number of files and rigorous checks, the paper argues that the geometry predicate defined as True invalidates the entire proof.
Rewritten by urgent.news; may contain errors — check the original.