OpenFrontiers
← Back to the index
OP-11979SolvedNumber theory

A Presentation of the Absolute Galois Group of Q2\mathbb{Q}_2

A Presentation of the Absolute Galois Group of Q2\mathbb{Q}_2. Solid result problem from Epoch AI’s FrontierMath: Open Problems; answers are checked by a verifier program.

Epoch field label “Number theory” mapped to MSC2020 11S20 for A Presentation of the Absolute Galois Group of $\mathbb{Q}_2$. https://epoch.ai/frontiermath/open-problems/q2-absolute-galois Not an author-supplied code.

Licensed source statementAs of Epoch AI data archive updated 2026-10-07; status as labeled by EpochSource & evidence ↓

Problem statement

The absolute Galois group of a field KK is the projective limit of all of the finite Galois groups Gal⁡(E/K)\operatorname{Gal}(E/K). It packages together the information about all finite extensions, and studying the Galois group of the rational field Q\mathbb{Q} is a central problem in algebraic number theory. One method for approaching this group is to study the analogous Galois groups of p-adic fields Qp\mathbb{Q}_p. In this case, there is an explicit presentation of Gal⁡(Q‾p/Qp)\operatorname{Gal}(\overline{\mathbb{Q}}_p/\mathbb{Q}_p) for p>2p>2. For p=2p=2, we have a description for the absolute Galois group of some extensions of Q2\mathbb{Q}_2, but not for Q2\mathbb{Q}_2 itself. Finding such a description would fill in a gap in our explicit understanding of Galois groups, and would have consequences in terms of counting p-adic fields with a given Galois group.

Read more here.

Progress Update (2026-07-06): Multiple AI systems have now generated solutions that are accepted by the verifier for this problem. The first solution we were made aware of was generated by Claude Fable 5, elicited by problem contributor David Roe on 2026-06-09. Then, on 2026-06-24, David Turturean wrote in with a solution generated by GPT-5.5 Pro running in a harness he designed. We believe that these solutions are likely to be correct, though it remains to check that the AI systems can actually prove this. In general FrontierMath: Open Problems verifiers provide strong evidence that an AI system has made meaningful progress on a problem, but for some problems, including this one, the evidence they provide is short of a full proof. In any case, since Fable was unavailable, Roe and Turturean began working to understand whether GPT-5.5 Pro could produce a full proof for its solution. We will provide an update here when they reach a conclusion.

Lean 4

Not yet attached

No Lean statement or proof is attached yet. Have a formalization? Share its source to add it to this problem.

Background

FrontierMath: Open Problems lists this problem with notability “Solid result”, field “Number theory” and type “Construction - Finite”. Each problem has a bespoke verifier program, so a proposed answer can be checked automatically.

Verifiable task — Epoch prompt, verbatim

Find a presentation for the absolute Galois group of the field of 2-adic numbers.  It may be helpful to know that the maximal pro-2 quotient of the absolute Galois group can be presented with three generators x,y,z and one relation, x^2*y^3*y^z=1.  The theory of Demuskin groups may also be useful.

You should give your output as two or three sections, separated by blank lines: variables, definitions (optional) and relations.

The variables section should consist of a comma separated list of variable names on one line corresponding to the generators of the presentation.  The first and second generators (say $\sigma$ and $\tau$ respectively) are distinguished, and must satisfy the implicit relation $\tau^\sigma = \tau^2$ (which should not be included in the \texttt{relations} section).  All later generators should be contained within the $2$-core of the group (the intersection of all $2$-Sylow subgroups).

In the definitions section, you may define additional variables in terms of the generators and variables defined earlier within the definitions section.  Each line should take the form "var = expr", where expr is an expression built using group operations.

In the relations section, you should give the relations as expressions in terms of variables defined in the previous two sections.  If there are multiple relations, give one on each line.

Expressions in profinite groups may involve exponents lying in $\hat{\Z}$.  To specify such a constant, you should provide its reduction modulo $85667662080 = 2^8 \cdot 3^2 \cdot 5 \cdot 7 \cdot 11 \cdot 13 \cdot 17 \cdot 19 \cdot 23$.

Example format for the solution (this is a presentation for p=3):

sigma,tau,x0,x1

tau2 = tau^3011753745
sigma2 = sigma^3011753745
rho = tau2^4
angle = (x0*tau*x0^-1*tau)^2974571600
u = (x1*(rho^2)*x1*(rho^2))^2974571600
v = sigma2*tau2
w = (u*(v^2)*u^-1*(v^2))^2974571600
y1 = x1^rho*(u^v)*(w^(sigma2*tau2^2))*(w^(tau2^2))
comm = x1^-1*y1^-1*x1*y1

(x0^sigma)^-1*angle*x1^3*comm

Progress & source assessment

Epoch marks this problem solved (AI) on 2026-06-09. This is Epoch’s classification; the solution has not been independently checked for this record.

Epoch solution update

First solved by: Claude Fable 5

This problem has been solved! See this write-up, by David Roe and David Turturean, for details.

References
  1. [1]Epoch AI, ‘FrontierMath: Open Problems’. Published online at epoch.ai. Retrieved from https://epoch.ai/frontiermath/open-problems (data updated 7 October 2026). — A Presentation of the Absolute Galois Group of $\mathbb{Q}_2$ linkCC BY 4.0 (Epoch AI data and problem pages); attributed reproduction of problem descriptions and verbatim task prompts.

Recorded progress

Add yours →

Nothing recorded yet. A pointer to a paper the entry is missing counts.

Machine attempts

0 recorded

Attempts run against the statement below and are recorded in full, including the failures. The point is to stop the next model — and the next person — from re-walking a road that is already known to end. Each attempt costs 25 credits.

No machine attempts recorded.

Sign in to run an attempt. New accounts start with enough credits for ten.

Discussion

0 notes

Notes are for clarifying the statement, flagging a missing reference, or recording why an approach fails. Claimed solutions belong in a submission, where they can be reviewed.

No notes yet.

Sign in to add a note.