Lean Verification of the Riemann Hypothesis - Joint statement by...
A proof of the Riemann Hypothesis is a major mathematical claim, and Alpöge released his in an unusually explicit form: a Lean library. On November 20, 2026, the Lean and Mathlib communities jointly announced that they had checked it and had full confidence in its correctness. Their certification matters because it rests on several distinct checks, not just a reading of the written argument. For example, the community confirmed that the library compiles and proves the same Riemann Hypothesis statement already recorded in Mathlib. Two independent tools compared the theorem in Alpöge’s library with Mathlib’s statement and found them to match. The exported proof was also checked by Lean’s official kernel and by independent proof checkers, including checkers with formal guarantees relative to set theory. The library contains 17 million lines of code, and dozens of people reported reproducing its compilation and raw proof export. The statement says the proof was checked on an inert export, outside the reach of hypothetical malicious code in the sources. The communities look forward to mathematical work explaining the proof’s new ideas.
What did Levent Alpöge release, and what did the Lean and Mathlib communities certify about it?
A proof of the Riemann Hypothesis is a major mathematical claim, and Alpöge released his in an unusually explicit form: a Lean library. On November 20, 2026, the Lean and Mathlib communities jointly announced that they had checked it and had full confidence in its correctness. Their certification matters because it rests on several distinct checks, not just a reading of the written argument.
For example, the community confirmed that the library compiles and proves the same Riemann Hypothesis statement already recorded in Mathlib. Two independent tools compared the theorem in Alpöge’s library with Mathlib’s statement and found them to match. The exported proof was also checked by Lean’s official kernel and by independent proof checkers, including checkers with formal guarantees relative to set theory.
The library contains 17 million lines of code, and dozens of people reported reproducing its compilation and raw proof export. The statement says the proof was checked on an inert export, outside the reach of hypothetical malicious code in the sources. The communities look forward to mathematical work explaining the proof’s new ideas.
How large is the Lean library containing the proof?
The source gives a striking scale: Alpöge’s Lean library comprises 17 million lines of code. A library is a collection of formal definitions and proof material that a system can process. In this case, the codebase depends directly only on Mathlib, Lean’s mathematical library. The figure describes the formalized library, rather than the length of a conventional prose explanation.
The article gives a concrete measure of verification: dozens of people reported that the codebase compiled correctly and reproduced the same raw proof export file. The statement was also isolated in a file importing only Mathlib, then compared with the theorem in Alpöge’s library using two independent tools. These checks concern the theorem and its formal proof, not merely the displayed line count.
The communities report that the library compiles without errors or build failures, and that independent proof checkers validated its exported proof. The source does not explain why the library has this many lines or how its size compares with other formal proofs. It does make clear that the scale was accompanied by coordinated checking, and that mathematical work explaining the proof’s ideas is still anticipated.
What does the Riemann Hypothesis claim about the zeros of the zeta function?
The Riemann Hypothesis is a claim about where certain zeros of the Riemann zeta function lie. In standard mathematical terms, it says that every nontrivial zero has real part one-half. The source article does not give this definition; it treats the hypothesis as a theorem whose statement is already present in Mathlib and whose proof has been formalized.
A zero is an input at which a function’s value is zero. The zeta function can be considered at complex inputs, which have both a real part and an imaginary part. The hypothesis singles out the nontrivial zeros and says that each sits on the vertical line with real coordinate one-half. The statement is not that every zero of every kind lies there.
This definition is standard mathematical background, rather than a detail supplied by the article. The article’s focus is instead on whether Alpöge’s Lean library proves exactly the Riemann Hypothesis statement in Mathlib and whether the proof checks. The Lean and Mathlib communities say both are correct, after compilation, statement comparison, and checks by multiple proof checkers.
What does it mean to express a mathematical proof in Lean rather than only write it in ordinary mathematical language?
A mathematical proof written in ordinary language relies on readers to interpret definitions, follow reasoning, and notice any gaps. A Lean proof encodes the statement and reasoning in a formal language with rules that software can check. This is general background about Lean; the article itself focuses on the checks applied to Alpöge’s formal proof of the Riemann Hypothesis.
In this case, the community reported that Alpöge’s library compiled correctly and produced a reproducible raw proof export. The exported proof was checked by Lean’s official kernel and by independent checkers, including con-leche and con-ron. The article says those two checkers have formal proofs, relative to set theory, that they cannot accept a false proof. This provides a second kind of assurance beyond successful compilation.
The communities also checked that the theorem in the library exactly matched Mathlib’s Riemann Hypothesis statement. They ran the checks on an inert export of the proof, outside the reach of hypothetical malicious code in the Lean sources. The article describes this formal verification as successful and says future mathematical activity will help explain the proof’s new ideas.
Why did the community compare the theorem in Alpöge’s library with Mathlib’s statement of the Riemann Hypothesis?
A proof can be valid yet still address the wrong claim if its formal statement differs from the intended theorem. That makes statement matching a crucial part of verifying a formal proof. The article says the Riemann Hypothesis statement in Mathlib had been carefully vetted by experts in Lean and number theory, so it provided a trusted reference for the comparison.
The community added a file that isolated the Riemann Hypothesis statement and imported only Mathlib. They then used two independent tools, comparator and con-leche, to compare that statement with the theorem proved in Alpöge’s library. Both tools confirmed that the theorem was exactly Mathlib’s statement of the Riemann Hypothesis, rather than merely a similar-looking claim.
This comparison addressed one of the article’s four verification criteria: that the code proves the correct statement. It complemented the separate evidence that the library compiled and that its exported proof passed multiple checkers. Together, these checks support the communities’ certification. The article reports that the theorem statement and proof are correct, while noting that further mathematical activity is expected to explain the proof’s ideas.
What does checking an inert proof export with Lean’s kernel and independent proof checkers help guard against?
A proof checker is meant to verify the proof itself, not trust arbitrary behavior in the source files around it. The article says the checks were run on an inert export of Alpöge’s proof, out of reach of hypothetical malicious code in the Lean proof sources. This helps separate verification of the formal proof from any concern that the source environment could interfere with the check.
The exported raw proof was checked by Lean’s official kernel, con-leche, con-ron, and several third-party checkers implemented in different programming languages. The article highlights that con-leche and con-ron come with formal proofs, relative to set theory, that they cannot accept a false proof. That addresses a separate concern: whether Lean’s logic itself might have flaws that a proof could exploit.
These safeguards do not replace one another. The community also checked that the formal statement matched Mathlib’s Riemann Hypothesis and that the library compiled. The article reports that all four criteria were satisfied and certifies the proof as correct. Its approach shows how an inert export and independent checking can strengthen confidence in a formal result.
Why is the Riemann Hypothesis important for understanding how prime numbers are distributed?
Prime numbers become less frequent as numbers grow, but their exact distribution is irregular. The Riemann Hypothesis is important because, in standard mathematical background, the zeros of the zeta function are closely connected to how that irregularity behaves. The article does not explain this connection; it focuses on verifying a proof of the hypothesis.
A useful way to understand the link is that the zeta function encodes information about primes, while its zeros influence estimates for prime-counting patterns. If the hypothesis is true, it implies strong bounds on the error in estimates of how many primes lie below a given number. This does not make primes fall into a simple repeating pattern, but it would sharply constrain their large-scale fluctuations.
The article reports that Alpöge’s Lean library proves the Riemann Hypothesis and that the Lean and Mathlib communities certified the statement and proof as correct. It does not discuss consequences for prime distribution or offer a mathematical explanation of the proof’s ideas. The communities say they look forward to activity that will help explain those new ideas.
This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.
Read more in the JupiteX app
Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.
Or read more news on the web