Net Good IndexSubmit a correction

Benefit Ledger · provisional · Mathematics

Claude Fable 5 produces a presentation of Gal(Q̄₂/Q₂)

On 9 June 2026 David Roe elicited a Claude Fable 5 solution to Epoch’s open problem asking for a presentation of the absolute Galois group of the 2-adics. Epoch later accepted verifier output and noted a second, independent GPT-5.5 Pro solution. Epoch warns the verifier is short of a full proof on this problem.

9 Jun 2026Tier 2 NotableMethodology 0.1

Current score

+0.26

3 base · Notable (tier 2 of 5, 3 pts)
× 0.7500 attribution · Critical contribution
× 0.4000 evidence · External expert evaluation
× 0.4500 realization · Experimentally validated
× 0.6500 durability
Event-level product before credit split: 0.26

An explicit Q₂ Galois presentation is notable (tier 2) until a complete proof is written. Fable 5 was first (0.75). Evidence is Epoch’s verifier plus a later writeup (0.40). Realization is a candidate presentation, not a textbook theorem (0.45).

What happened

Presentations of Gal(Q̄_p/Q_p) are classical for odd p; p=2 was a documented gap. Epoch lists Fable 5 as first solver, with a later GPT-5.5 Pro writeup by David Turturean. Because Epoch says verifier evidence here is short of a full proof, the event is provisional Notable rather than Major, and Fable 5 keeps first-solver credit rather than splitting equally.

Model attribution

Claude

First system Epoch records as solving the Q₂ absolute Galois presentation problem, elicited by David Roe.

Epoch names Claude Fable 5 as first solved; a later GPT-5.5 Pro solution is independent confirmation, not a co-search on the original run.

Attribution 0.7500 · Credit share 100% · Anthropic

Claims

  • Epoch records a Claude Fable 5 solution, elicited 9 June 2026, to a presentation of the absolute Galois group of Q₂.

    outcome · supported

  • Epoch’s verifier constitutes a complete proof of the presentation.

    significance · disputed

Sources

primary sources

Revision history

  • 13 Sep 2026 · 0.00 0.26

    Imported events under methodology 0.1.