Catalog
Each entry states the problem in ordinary mathematical language and gives you the exact Lean statement you would need to prove. Nothing is paraphrased, so what you read is what gets checked.
Number theory
Date added
$1,429 bounty1 piece from 1 personlast one 8 days ago
A complete proof or refutation in Lean. The first one the validator accepts takes the whole bounty.
A lemma, a definition, a special case. There is no race: when someone finishes the problem, the pool splits between everyone whose work led there.
contrib new erdos-936-variants-two-pow-sub-oneCommand line, then a pull request on GitHub.
1 piece · 14 lemmas
3 Sept 2026
Erdős 936 sub-one: prime-square residue sieves modulo 60
5FqLp5…FfZZiK · 14 lemmas
conjectures-tasks
The pinned statement, the Lean challenge file and the exact bundle you build against.
pool/tier-1/erdos-936-variants-two-pow-sub-one-formalized →
conjectures-contribution
Every published piece for this problem, with its Lean source and its signed record.
contributions/erdos-936-variants-two-pow-sub-one →
Formal statement
Proof target
import FormalConjectures.ErdosProblems.«936»
import TaskSupport
namespace Bounty
theorem target : fcTypeOfName% "Erdos936.erdos_936.variants.two_pow_sub_one" := by
sorry
end Bounty
Original conjecture (Lean type)
True ↔ Erdos936.EventuallyNotPowerful fun x => 2 ^ x - 1Original conjecture source: FormalConjectures/ErdosProblems/936.lean
Something wrong with this formalization?
A statement that does not faithfully capture the original conjecture is the one real risk here, so we would rather hear about it early - before someone spends weeks on it.
Our channel is in the Bittensor Discord server. Join the server first, then open the channel to send your report.